Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
A study of zwitterionic adducts of TCNQ.
BROUGHTON, Richard A.
Available from the Sheffield Hallam University Research Archive (SHURA) at:
http://shura.shu.ac.uk/19400/
A Sheffield Hallam University thesis
This thesis is protected by copyright which belongs to the author.
The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the author.
When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given.
Please visit http://shura.shu.ac.uk/19400/ and http://shura.shu.ac.uk/information.html for further details about copyright and re-use permissions.
2
Sheffield Hallam University
r e f e r e n c e o n l y
ProQuest Number: 10694281
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com p le te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
uestProQuest 10694281
Published by ProQuest LLC(2017). Copyright of the Dissertation is held by the Author.
All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode
Microform Edition © ProQuest LLC.
ProQuest LLC.789 East Eisenhower Parkway
P.O. Box 1346 Ann Arbor, Ml 48106- 1346
A STUDY OF ZWITTERIONIC ADDUCTS OF TCNQ
by
RICHARD ANTHONY BROUGHTON BSc
A thesis submitted to Sheffield Hallam University in partial fulfilment of the requirements for the Degree of Doctor of Philosophy.
Sponsoring Establishment: Division of Applied ChemistrySheffield Hallam University
Collaborating Establishment: Health & Safety Executive
February 1993
CONTENTS
Synopsis
CHAPTER 1: ELECTROACTIVE ORGANIC COMPOUNDS
1.1 Historical Introduction 1
1.2 Organic Charge-Transfer Complexes 4
1.3 Conductive Organic Charge-Transfer Complexes 6
1.3.1 Organic Metals 7
1.3.2 Organic Superconductors 12
1.3.3 Some Physical Concepts 14
1.4 Pyridinium and Quinolinium TCNQ Compounds 151.5 Organic Materials for Non-Linear Optics - Optical Second Harmonic
Generation23
1.6 Moleecular Rectification 27
1.7 References 30
CHAPTER 2: SYNTHESIS AND CHARACTERISATION OF R(4)Q3CNQ/R(4)P3CNQ
2.1 Experimental 34
2.1.1 Reagents 34
2.1.2 Microanalysis 34
2.1.3 Instrumentation 34
2.1.3.1 Infra Red Spectroscopy 34
2.1.3.2 Ultra Violet/Visible Spectroscopy 34
2.1.3.3 Mass Spectroscopy 36
2.1.3.4 ]H nmr Spectroscopy 36
2.1.3.5 Differential Scanning Calorimetry 36
2.2 Synthesis 36
2.2.1 Synthesis of the N-alkyl-4-methyl quinolinium bromides 36
2.2.2 Synthesis of C20H41(4)Q3CNQ 37
2.2.3 Synthesis of LiTCNQ 39
2.2.4 Synthesis of Me(4)P3CNQ 39
2.3 Discussion of Synthetic Methods 41
2.4 Spectroscopic Studies of R(4)P3CNQ/R(4)Q3CNQ 51
2.4.1 Ultra Violet/Visible Spectroscopy 51
2.4.1.1 Solvatochromism in R(4)P3CNQ/R(4)Q3CNQ 61
2.4.2 Infra Red Spectroscopy Studies 65
2.4.3 Mass Spectroscopy Studies 69
2.5 Miscellaneous Studies 84
2.6 References 85
CHAPTER 3: LANGMUIR-BLODGETT FILMS
3.1 Historical Introduction 86
3.2 Molecular Requirements of LB Film Forming Materials 88
3.3 Isotherm Measurement 93
3.4 Deposition and Mono/Multilayer Assembly 98
3.5 Characterisation of LB Films 101
3.6 Applications of LB Films 104
3.7 References
CHAPTER 4: LANGMUIR-BLODGETT FILMS - EXPERIMENTAL
4.1 The Joyce Loebl Monolayer Coating Unit 111
4.2 The Compression System 111
4.3 Surface Pressure Measurement 113
4.4 Trough Operation 117
4.4.1 Cleanliness 117
4.4.2 Trough Calibration 118
4.4.2.1 Calibration of Surface Area 118
4.4.2.2 Calibration of Surface Pressure 118
4.5 The Subphase 119
4.6 Subphase Surface Cleaning 119
4.7 Monolayer Material Preparation and Spreading 120
4.8 Temperature of Subphase 121
4.9 Compression of the Monolayer 121
4.10 Control of Surface Pressure 122
4.11 Transfer of Monolayer 122
4.12 References 126
CHAPTER 5: LANGMUIR-BLODGETT FILMS - RESULTS AND DISCUSSION
5.1 Langmuir Film Studies on R(4)Q3CNQ 127
5.2 Langmuir-Blodgett Films of R(4)Q3CNQ 144
5.3 Characterisation of LB Films of R(4)Q3CNQ 145
5.4 General Discussion 164
5.5 Change in Film Structure and Absorbance Bands with Time 188
5.6 References
Acknowledgements
SYNOPSIS
Chapter 1 gives a general introduction to the field of electroactive organic compounds. The
historical development of the field is outlined and then the more specific areas are described in
detail. The properties of picolinium and quinolinium TCNQ salts are introduced and the
extension of this to the molecules studied in this work developed. Two current areas of
interest - second harmonic generation and molecular rectification - are then reviewed and the
potential applications of the materials studied in this work highlighted.
Chapter 2 discusses the synthesis of the materials and contrasts the two methods used. A
discussion of the reaction mechanisms is given, as well as techniques for monitoring the course
of reaction. The characterisation of the materials by a range of spectroscopic techniques is
described. The solvatochromic behaviour of the materials is shown to conform with theoretical
predictions and the observed mass spectra are shown to have some diagnostic importance.
4
Chapter 3 gives a general introduction to Langmuir-Blodgett (LB) films starting once again
from a historical perspective. The molecular requirements of suitable materials together with
pressure/area per molecule measurements and the deposition process are described. The
various methods available for the characterisation of LB films are described, as well as the
many potential applications proposed within the general field of molecular electronics.
Chapter 4 describes the experimental methods employed when using the Joyce Loebl Langmuir
Trough. The importance of parameters such as cleanliness, sample purity and instrument
calibration is stressed.
/Chapter 5 discusses the behaviour of the materials on the subphase and their resultant
fabrication as LB films. The structure of the films is shown to be dependent on the
hydrophobic chain length with a definite change in film structure occurring at a 15 carbon
chain. Reasons for this are proposed and calculations of chromophore/hydrocarbon chain tilt
angles on the basis of the proposed structure are given. A time dependent change in film
structure is also discussed, as well as suggestions for future work.
CHAPTER 1: ELECTROACTIVE ORGANIC COMPOUNDS
1.1 Historical Introduction
The electrical conductivity of most organic materials, when purified, is extremely
low ( a n < 10“ 10 Scm-1) at room temperature. It was suggested at the turn of the
century1' 2 that organic solids may exhibit conductivities comparable with metals.
Indeed, today there are two classes of organic solids - polymeric hydrocarbons and
charge-transfer complexes - of which certain members may be referred to as
organic metals.
Strong 7r-molecular donor (D) and acceptor (A) molecules can react to produce ion-
radicals salts (X.+A'. or D.+X*.), where 'X ' is an appropriate anion or cation, or charge-
transfer compounds (DA and D+A*.). Such compounds may in some instances exhibit
higher conductivities than is normally found in organic materials. A stable charge-
transfer complex is formed, as the name implies, by the transfer of an electron from
a donor to an acceptor to produce an ionic crystal in which both the anion and the
cation are complex chemical units in their own right. The more commonly
encountered donors are amines, electron rich alkenes, heterocycles and simple alkali
metals. Common electron acceptors include quinones, electron deficient alkenes
and heterocycles and the halogen atoms. Examination of the crystal structure of
such compounds show that the donor and acceptor molecules stack face-to-face
alternately (fig 1.1).
1
D A D A
A D A D
D A D A
A D A D
Fig 1.1 Mixed donor-acceptor stacks
For a more detailed comprehensive review of organic charge- transfer complexes
the reader is referred to the excellent text by Foster3. In the early 1960’s, Melby
and his co-workers4 at duPont reported the synthesis of 7,7,8,8 - tetracyano-p-
quinodimethane (TCNQ -fig 1.2), a new, powerful organic electron acceptor.
Fig 1.2 7,7,8,8 - tetracyano-p-quinodimethane (TCNQ)
A number of complexes were produced between TCNQ and various donors and
many were found to be semi conductors ( o n « 10“5 Scm"1), and indeed the
quinolinium-TCNQ complex was the best organic conductor known at the time.
This large improvement in conductivity prompted renewed interest in this field,
focusing primarily on TCNQ salts.5 It was soon realised that high conductivity was
associated with segregated stacks of donor and acceptor molecules once again
packing face-to-face (fig 1.3). The resulting charge transfer and x-overlap between
molecules is very strong and thus unpaired electrons are partially delocalised along
2
the one dimensional stacks. The resultant conduction is therefore highly anisotopic.
D A D A
D A D A
D A D A
D A D A
Fig 1.3 Segregated donor and acceptor stacks
In 1970,6 a powerful new donor, tetrathiofulvalene (TTF) (fig 1.4) was synthesised
and three years later its TCNQ salt was prepared7 which was found to be highly
conducting at room temperature - (crr t « 500 ScrrT 1)
Fig 1.4 Tetrathiofulvalene (TTF)
This was the first organic metal - a dramatic increase in conductivity occurring
below room temperature, rising as high as 104 ScnrT1 at 60K. Below this
temperature, a metal-insulator transition occurs, which can be controlled to some
extent by the chemical structure.
3
During this period there was also extensive interest in polymeric hydrocarbons,
with extensive x-electron delocalisation leading to a weak conduction band.
Much attention has been focused on a polyacetylene which in its ' pure state' is an
insulator (a t ~ 1CT9 Scrn”1) due to a relatively large band gap. Heeger and co-
workers8 showed that doping with strong reducing or oxidising agents can increase
the conductivity dramatically ( a t « 500 Scrn-1 , for example). Two other classes
of one dimensional conductors include polymeric materials of the main group
elements, eg (SN)9 and linear chain transition metal compounds10 where intra chain
overlap may include the ligand x-system. The remainder of this introduction will
not concern itself with polymeric materials any further; rather it will concentrate
upon organic charge transfer salts based on TCNQ.
1.2 Organic charge - transfer complexes3
The complex formed between iodine and benzene11 was one of the earliest charge
transfer complexes on which much work was carried out. As is typical of such
complexes, the ultraviolet/visible (uv/vis) spectrum showed features characteristic
of neither the solute nor the solvent. Rather, features characteristic of some form
of complex between the two was observed. The explanation of this observation led
to an extension of Lewis acid-base theory and thus explanations of many of the
phenomena associated with molecular complexes were forthcoming; for example,
the very intense absorption bands in the uv/vis region of the spectrum.
In the early 1950's,12,13 Mulliken published a series of papers in which he
described electron donor-acceptor complexes. By definition, the donor (D)
possesses a HOMO (Highest Occupied Molecular Orbital) and the acceptor (A) a
LUMO (Lowest Unoccupied Molecular Orbital). From this, therefore, charge
4
transfer occurs between the HOMO of the donor and the LUMO of the acceptor, as
the HOMO is of high energy and the LUMO low. The bonding between the D and
A was described14 by the following wave function:
^ N (A.D) = a f 0 (A,D) + b ^ 1 (A -D +) Eq (1)
where 'a' and 'b' are small integers, and a>b; Q is the "no-bond" contribution
and a dative bonding contribution. T hus,^ (A.D) corresponds to the wave
function which adequately describes the resultant combination of two extreme
resonance forms. Eql also shows that, while the degree to which charge transfer
occurs may influence the nature of the complex, it does not dictate the bonding
within that complex.
For any electron transfer to take place, the energy levels of the donor and acceptor
need to be matched. In solution, this can be achieved by the reorientation of the
solvent around both the donor and acceptor, and in the solid state by the
reorientation and rotation of bonds. Thus, the energy barrier to electron transfer is
the summation of these effects.
The energy of the transition from the donor to the acceptor is dependent on the
ionisation energy of the donor (ID) and the electron affinity of the acceptor (EA).
Very simply, if a strong donor is in close proximity to a strong acceptor, the energy
of the charge transfer is small, the ground state is for all intents and purposes ionic
and essentially 100% charge transfer occurs. For weak donors and weak acceptors,
the converse is true - the energy of charge transfer is large, the ground state is
neutral and the resultant degree of charge transfer is essentially zero.
5
It has been shown that, if the energy of charge transfer is small (and positive), then
partial charge transfer occurs, and the resultant conductivity is high. The reason for
this shall be discussed in more detail in subsequent sections.
1.3 Conductive Organic Charge-Transfer Complexes15
Homomolecular organic materials have extremely low electrical conductivities. At
the very best they are wide band-gap semi conductors (cr t « 10"9 to 10"14 Scrn"1),
though the vast majority are insulators. As was mentioned earlier, the idea of
highly conductive organic materials was first suggested many years ago, but only
recently has research effort been rewarded in this field. This was partly due to
development of the solid state techniques used to study inorganic semiconductors.
The realisation that the highly conductive materials were generally charge transfer
complexes, coupled with the synthesis of TCNQ at duPont led to an upsurge in
interest throughout the 1960's and 70 's.
TCNQ is prepared, in two stages, by the condensation of cyclohexane -1,4-dione
with malonitrile followed by oxidation in a single step (fig 1.5).
TCNQ readily undergoes a one electron reduction and the resultant anion forms a
variety of charge transfer complexes and radical anion salts with suitable donors.16
For example, stable complexes are produced with quaternary quinolinium and
picolinium cations, and these shall be discussed in more detail later. A wealth of
structural data has been obtained on TCNQ complexes, and the conductive, charge-
transfer species have been shown to crystallise in segregated stacks with the
conductivity greatest in the stacking direction. In this structure, the predominant
overlap is between molecules of the same type and thus orbitals of similar energy.
6
0
Fig 1.5 The 2-step synthesis of TCNQ
This structure is much more efficient at producing bands than is overlap between
different molecules. With the realisation that the TTF - TCNQ complex exhibited
metallic conductivity? much work focused on the search for new organic metals and
an understanding of their physical properties.
1.3.1 Organic Metals17
The TTF-TCNQ system has been extensively studied for a number of reasons
including:
i) it is relatively straight-forward to grow good-sized, robust, single crystals.
7
ii) its properties are essentially characteristic of the class of organic metals as a
whole.
As has already been mentioned, the crystal structure of TTF-TCNQ showed
segregated stacks of donor and acceptor molecules, uniformly spaced and
interlocking18(fig 1.6). It can be seen that, within such stacks, the donor and
acceptor molecules do not sit directly on top of each other, rather there is a
sideways displacement which leads to a "ring-double bond overlap", in which the
exocyclic C =C bond lies above the ring system of the molecule next to it in the
stack.
On cooling, the conductivity of the system steadily increases from 500 Scrn-1 at
293K to in excess of 104 Scrn-1 at 59K. However, on cooling the system further,
three successive phase changes occur, at 59K, 47K and 38K. Eventually, an
insulating state is produced with three dimensional order. The many studies that
have been carried out on the system interpret the phase transition as a Pierels
distortion19' 20. This is expanded on in Section 1.3.3, though essentially, the
Pierels theorem states that any one-dimensional system will become unstable with
respect to lattice distortions and long-range order cannot be maintained.
Importantly, the period of such lattice distortions is related to the number of
electrons in the band and calculations based on this show that partial charge transfer
occurs from donor to acceptor, resulting in 0.59 electrons in the TCNQ bands.
Hence, both stacks are partially filled and both stacks can contribute to the
conductivity.
8
Figure 1.6 The crystal structure of TTF-TCNQ (from ref 18)
The segregated stacks and "ring-double bond" overlap
can be seen
9
It was soon clear that TTF and TCNQ had certain properties that made them unique
amongst the other donor and acceptor molecules available at the time. The
ionisation potential and electron affinities were very similar, a condition which was
previously shown to favour partial charge transfer and consequent high electrical
conductivity. They are planar and, with point group D2h, have a high degree of
symmetry. The t -delocalisation extends throughout the molecule and both are of
similar size.
Tremendous effort has gone into the synthesis of new donors with a view to testing
the applicability of these factors and also with the ultimate aim of stabilising the
metallic state and producing super-conductivity. Most of the synthetic work has
concentrated on the search for derivatives of the 1,3-dithiole ring system of TTF.
Examples are shown in fig 1.7.
Notice there has been extension of both the a 21 and x-bond22 framework.
Successful replacement of the sulphur atoms by selenium atoms23 and tellurium
atoms24 has also been achieved.
The synthesis of new acceptors has been the subject of very much fewer
publications, presumably because the chemistry of producing new TCNQ
derivatives is very demanding. However, a series of three early papers 25~27
reported the synthesis of 21 derivatives and also the results of studies on some of
the complexes with TTF. From this work several general points can be made and
these can be illustrated by a comparison of two salts, HMTSF-TCNQ and HMTSF-
TCNQF4 (the structure of TCNQF4 is shown in figure 1.8, HMTSF is the selenium
analogue of HMTTF shown in fig 1.7).
10
Fig 1.7 Some derivatives of TTF
X K X
X Xx
X = S: Hexamethylenetetrathiofulvalene (HMTTF)
X = Te: Hexamethylenetetratellurafulvalene (HMTTeF)
Cyclohexa-2,5-diene-l,4-diylidene-bis-l,3-benzodithiole
HC
S e ^
S e
Tetramethyltetraselenafulvalene (TMTSF)
NC
Figure 1.8 TCNQF
HMTSF-TCNQ is an extremely conductive organic charge transfer complex ( a r t
~ 1500 Scrn-1) whilst HMTSF-TCNQF4 has a conductivity many orders of
magnitude smaller (ar t ~ 10-6 Scrn"1). The key to this lies in the degree of
charge transfer. In TCNQ-HMTSF there is partial charge transfer whilst in
TCNQF4-HMTSF, charge transfer is complete because TCNQF4 is a much
stronger electron acceptor due to the presence of the extra highly electronegative
fluorine atoms. This rule has found to be quite general; substitution of the quinoid
ring of TCNQ reduces the conductivity of any resulting complex relative to TCNQ
itself, with the reduction being particularly marked if the substitution produces a
stronger electron acceptor.
In such organic conductors the substitution of selenium for sulphur stabilises the
metallic state and the donor stack dominates the transport properties due to the
presence of ' d'-orbitals.
1.3.2 Organic Superconductors
In 1978, the complex TMTSF-2,5 - dimethyl TCNQ was found to conduct
electricity under pressure down to IK; at ambient pressure the now familiar metal-
insulator transition occurs at 42K28 ,29. It was proposed that the conductive state
may be stabilised due to the superconducting pairing of electron states. This theory
promoted the synthesis of many cation radical salts of the form (TMTSF)2 X where
12
'X ' is an inorganic anion such as PF6. In 1980 it was shown by Bechgaard30 that
(TMTSF)2 PFfi exhibited superconductivity at IK under a pressure of 12000
atmospheres. Indeed, many members of the (TMTSF)2 X series exhibit
superconductivity, with, for example, (TMTSF)2 C104 showing superconductivity
at ambient pressure down to 1.4K.
The theory of metallic superconductivity [the well known BCS theory] was first
formulated in 195731 and was based on the ordered motion of electron pairs -
Cooper pairs - and not individual valence electrons. Evidence now suggests that for
salts of the (TMTSF)2 X family, Cooper pairing does indeed occur32. This pairing
is localised along the donor chains - often at temperatures up to 30K. This
temperature is well above the onset of bulk superconductivity.
One particularly striking feature of the (TMTSF)2 X salts is the zig-zag stacking
arrangement of TMTSF molecules into sheets which are then separated by anions.
There are also both inter and intra stack contacts of the selenium atoms33' 34, the
result of which is a large, two- dimensional sheet of molecules.
This is necessarily only a brief introduction to the field of organic superconductors
of which a wealth of data now exists. In recent years, this class of compounds has
really been superceded, to some extent, by the ceramic superconductors. For
example, claims of superconductivity at temperatures as high as 60K have been
made for La2_xSrxCaCu20 635.
13
1.3.3 Some Physical Concepts
The structure of highly conducting organic charge-transfer complexes shown in fig
1.3 leads to the proposal that such complexes are one- dimensional molecular
arrays. They consequently have more than their complement of valence electrons
required for bonding.
Theoretically, it can be considered that such extra electrons partially fill a
conduction band where the interaction between neighbours determines the
bandwidth. Pierels?-9 in a classic text, pointed out that such a "quasi" one
dimensional metallic system could not maintain long range order at low
temperatures and would become unstable with respect to lattice distortions. The
period of the lattice distortion can be related to the number of electrons in the band.
Very simply, the conducting chain becomes stretched in one region and contracted
in another, thus localising the conducting electrons. The bonding between atoms is
much stronger where they are most closely spaced, and the electrons have a lower
energy in these regions. There is consequently a localised concentration of
electrons along the chain known as a charge density wave which, if in phase with a
periodic lattice distortion, results in an energy gap and an insulating or, at best,
semi-conductive state. This is a very simple description of the Pierels Theorem.
However, charge-density wave-phonon interactions (ie charge-density waves
coupled to underlying lattice vibrations) could enable the wave to travel freely
through the crystal generating Frohlich Superconductivity in a mechanism different
to BCS pairing. If the number of electrons in the chain is commensurate with the
number of lattice sites, then a metal-insulator transition occurs due to the high
potential energy needed to be overcome to enable the electrons to move to new
equilibrium positions. If the number of electrons is incommensurate with the lattice
14
then they will be free to move, as the charge-density wave is not locked to the
lattice. Thus, we have a mechanism for high conductivity and for a truly
incommensurate system Frohlich superconductivity would occur.
Electron spin density distortion along a chain provide another possible instability for
one-dimensional conductors. Electron spin density may distort along the chain
producing in extreme cases an anti-ferromagnetic alignment of spins which can
preclude superconductivity. Interactions in one- or two-dimensions are suggested to
prevent the onset of spin density waves, particularly in an analysis of the Bechgaard
salts.
1.4 Pyridinium and Quinolinium TCNQ Compounds
The work documented in this thesis involves the synthesis and preliminary
characterisation of 7r-bonded pyridinium and quinolinium TCNQ systems. The
published literature on the properties and structure of radical anion salts of these
cations is now discussed.
Melby et al16 published the first examples of a complex between TCNQ and a
nitrogen heterocycle in 1962 - the complex produced when the lithium salt of
TCNQ was allowed to react with N-methylquinolinium iodide. Lithium TCNQ is
very soluble and is often used in the synthesis of TCNQ complexes for this reason.
Along with similar salts of pyridinium cations, the electrical conductivities were
very low ( a t * 10“7 Scrn"1).
The substitution of the pyridinium ring36 by various alkyl groups did not markedly
improve conductivity - these complexes had conductivities in the range 10“6 Scrn” 1
to 10“8 Scrn-1 .
15
The conduction path in such simple TCNQ salts is dependent on the excitation of an
electron from one ionic state to another.
TCNQ TCNQ -------------------- > TCNQ TCNQ2“
Notice that in this mechanism we are placing two conduction electrons on the same
TCNQ site.
An estimation of the band gap can be made by consideration of the Coulombic
repulsion energy (U) between electrons on adjacent sites.
So:
e2U = — Eq.(2)
47r£ do
Where 'e ' is the electronic charge; ' £ 1 the relative permittivity of free space and
'd ' the charge separation measured in X.
Thus: for two electrons
14.4U = eV Eq.(3)
d
The distance between exocyclic carbon atoms in TCNQ is 5.5X. The
intermolecular interplanar spacing is 3.3A and the intermolecular distance is 9X.
Thus for two electrons on the same site the repulsion energy is calculated as 2.6eV,
and for adjacent sites 1.6eV, if we assume that the conduction electrons localise on
the C(CN)2 groups at opposite ends of the molecule. Thus, the electrical
conductivity is low, as placing two electrons on the same TCNQ site is energetically
unfavourable.
16
More complex salts of TCNQ involve various molar proportions of neutral TCNQ
with TCNQ and the cation. For example:
M TCNQ”, (TCNQ)
M^ITCNQ- )2 (TCNQ)
The electrical conductivities of complexes of this type are many orders of
magnitude larger than those of simple salts, for example - 1CT4 and 10"1 Scrn"1 at
300K37. This is due to the presence of neutral TCNQ molecules within the chains.
In these cases the necessity of placing two conduction electrons on the same site is
removed and a corresponding increase in electrical conductivity is observed.
Although the conductivities of simple TCNQ salts is low the long chain N-
alkylpyridinium TCNQ salts have been studied as Langmuir-Blodgett films ,(see
Chapter 3) by Barraud and his co-workers38 in France. Their conductivity has
been reported, and the above material has been exposed to gases such as Cl2 and
NO, and its response to these discussed39.
Complex salts of quinolinium and pyridinium cations were originally prepared by
one of three methods:17
•f- *» 4- —1) C TCNQ + TCNQi±£!LC (TCNQ)2 + I2
2) 2C?"f + 4TCNQ McCN, 2C+ (TCNQ)2 + I24- —
3) 2D + H2TCNQ + 3TCNQ M‘CN>2DH (TCNQ)2
Where C is an organic, organometallic or inorganic cation and 'D ' is a nitrogen
heterocycle.
17
Addition and substitution reactions of TCNQ have been documented for many
years40. Substitution at either, or both, of the cyano groups occurs readily with
certain amines41 (fig 1.9).
n(X X n
nc ■v.N
_ Z _ - - - - - >
m C '
t - H C N lr2hh
mC cn
Fig 1.9 Substitution of cyano groups in TCNQ by secondary amines
18
The structure (1) can be represented by the two resonance structures shown in
figure 1.10.
The infra-red spectrum of this adduct shows two distinct nitrile stretching
frequencies at 2175 cm”1 and 2130 cm"1, characteristic of a monosubstituted
malonitrile anion. The uv/vis spectra of the same species in acetonitrile showed a
broad band which was attributed to an intramolecular charge transfer from the
donor to the acceptor. In 198442, the zwitterionic donor-x-acceptor adduct Z-
R
H C < CN
Fig 1.10 Zwitterionic canonical forms of amino adduct (1)
19
(N-methyl-2- pyridinium)-a-cyano-4-styryldicyanomethide (trivial name picolyl-
tricyanoquinodimethane) was synthesised in these laboratories in Sheffield. Its
structure is shown in fig 1.11 and its molecular geometry in fig 1.12. The bond
length between C12 and C15 clearly shows this is a double bond whilst the
benzenoid nature of the two rings is also confirmed by bond measurements.
\
Fig 1.11 Picolyl-tricyanoquinodimethane
The molecule is not planar, the pyridinium ring is twisted from the benzenoid ring
of TCNQ by 10.13°. The uv/vis spectrum, in acetonitrile, does not show the
typical TCNQ band characteristic of the anion radical; rather, a broad band is
present, centred on 592 nm. It has been suggested that this is an intramolecular
charge transfer transition. Fuller spectroscopic studies of this and related molecules
are discussed in more detail in Chapter 2.
20
Fig 1.12 Crystal structure of Picolyl-tricyanoquinodimethane
21
The charge distribution was shown to be typical of a zwitterion by atom- in
molecule calculations carried out on the structure. Evidence for such a zwitterionic
structure was also obtained from the very large computed dipole moment of
26 .16D. The molecule has also been studied in its solid state by polarised reflection
spectroscopy43. An intramolecular charge transition was observed at 537 nm. The
stacking of the molecules in the solid state is shown in fig 1.13, clearly showing a
head-to-tail arrangement.
Fig 1.13 Overlap of two picolyl-tricyanoquinodimethane molecules
stacked along the b-axis
22
1.5
Such a head-to-tail arrangement has recently been observed by Miller and
Calabrese44 in the zwitterionic adduct of tetrafluoro TCNQ and
tetramethylphenylenediamine,(TMPD) whose structure is shown in fig 1.14.
The structure of such zwitterionic donor-x-acceptor adducts opens the door on
potential applications in a number of exciting new fields and these are discussed.
Fig 1.14 The structure of the TMPD-TCNQ zwitterionic adduct
Organic Materials for Non-Linear Optics - Optical Second Harmonic
Generation45
Traditionally, optics and electronics have relied on inorganic compounds for their
various components, though in the future they will no doubt benefit from the vast
range of organic materials already known and the many more waiting to be
synthesised.
F F
23
Consider the change in dipole moment which occurs between the ground state /xg
and an excited state upon interaction of an electric field - say the electrical
component of electromagnetic radiation - with a single molecule. This can be
expressed as a power series of the electric field:
quadratic and cubic hyperpolarisibilities respectively. As the hyperpolarisibility
coefficients are tensor quantities, they are symmetry dependent and it can be shown
that odd-order coefficients are non-zero for all molecules, but even-order
coefficients, such as^are zero for centrosymmetric molecules. The quadratic term
is responsible for second harmonic generation (SHG). Note that equation (4) is, in
effect, identical to (5) except that (5) refers to the macroscopic polarizability such as
may occur in a crystal.
As only about 20% of organic molecules have a non-centro-symmetric structure46,
so molecules with large predicted values of fl may exhibit no SHG.
Eq(4)
or P = Po + X (1>E + X (2)EE + X (3 )EEE + Eq(5)
The coefficient ' a ' is the well known linear polarisibility, and are the
24
The Langmuir-Blodgett technique - which will be discussed in detail later - enables
us to orientate molecules in a non-centrosymmetric fashion and, indeed, SHG has
been observed for such a system47.
What, therefore are the structural features that would lead us to predict a frequency
doubling effect? Consider the selection of molecules shown in table 1. The
bracketed numbers give the enhancements in efficiency one might expect if these
molecules were optimally aligned in a crystal. The object of the table is not to give
any hard and fast data on SHG in organic materials, but just to enable certain trends
and points to be noted.
First, active molecules are conjugated and secondly they are polarised, ie they
contain both donor and acceptor substituents. Obviously, the presence of a
conjugated and polarised system is not mandatory, though materials which are the
most efficient appear to exhibit these two criteria.
Thus, on the basis of their determined structure the highly conjugated, donor-
acceptor zwitterionic adducts studied in this work may exhibit second harmonic
generation if suitably aligned.
25
TABLE 1
TYPICAL ORGANIC MATERIALS FOR OPTICAL SECOND HARMONIC
GENERATION (from Ref 54)
Structure Efficiency fi, (xlO
hi 6
0 I
1.0(x2.5) 0.45
0 I
'] 0.0
c h 3
J L 13(x4)
30esu)
1.6 Molecular Rectification
The construction of a simple, electronic rectifier based on the use of a single
organic molecule was discussed by Aviram and Ratner in 1974?5' 56
The most common of solid state rectifiers are based on a p-n junction and so for an
organic molecule to show rectification properties it must also have the properties of
a p-n junction. By altering the substituents on an aromatic system within the
organic moiety, it is possible to increase or decrease the electron density in the
molecules, thus making the species more electron rich (n-type) or electron poor (p-
type). Any electron withdrawing subunits cause an aromatic system to become
electron poor in 7r-density, raising the electron affinity and making the species
a good electron acceptor. Conversely, any electron releasing groups increase tc-
electron density, consequently lowering the electron affinity and producing a good
electron donor.
This explanation lacks the rigorous treatment given in any solid state physics
course, but is used as a guide as to why the structures chosen for study as potential
molecular rectifiers consist of donor and acceptor species.
In a donor-acceptor system, electrons would pass from the cathode to the acceptor
and from the donor to the anode, but not in the other sense. Interaction of the
donor and acceptor must be minimised or a single donor level would exist for the
duration of any laboratory experiment.
To insulate a donor and acceptor from each other, a cr-electron bridge to link the .
two portions was proposed and the following molecule (fig 1.15) suggested as a
model.
27
C n
N
Fig 1.15 Proposed Molecular Rectifier based on the Aviram/Ratner model.
A successful synthesis of this molecule has not been achieved but Metzger et al57
have successfully synthesised a variety of molecules of similar structure (fig 1.16,
for example) and have recently reported partial success58
HN 0
Fig 1.16 TTF and a TCNQ derivative linked via a urethane unit
28
There have been a number of publications recently on the novel Zwitterionic
materials documented in this thesis. Ashwell and co-workers have reported on their
photochromism59'61 and possible molecular rectification62. Aspects of the LB film-
forming properties have recently been reported by Bell et al63. For current
developments in electro-active organic compounds in general, then the proceedings
of the many workshops, conferences, etc provide the necessary background64.
29
1.7 REFERENCES
1 H N McCoy and W C Moore, J Am Chem Soc, 33, 273, (1911)
2 C Kraus, J Am Chem Soc, 35, 1732, (1913)
3 R Foster, Organic Charge Transfer Complexes, Acadmic Press London &New York, 1969
4 D S Acker, R J Harder, W R Hertler, W Mahler, L R Melby, R E Benson and WE Mochel, J Am Chem Soc, 82 , 6403, (1960)
5 I F Shegolev, Phys Stat Sol, (A), 12, 9, (1972)
6 F Wudl, G M Smith and E J Hufnagel, J Chem Soc Chem Commun, 1453.(1970)
7 J Ferraris, D O Cowan, V V Walatka and J M Perstein, J Am Chem Soc, 95, 948,(1973)
8 H Shirikawa, E J Lewis, A G MacDiarmid, C K Chiang and A J Heeger, J ChemSoc Chem Commun, 578, (1977)
9 M M Labes, P Love and L F Nicholls, Chem Rev, 79, 1, (1979)
10 A E Underhill and D M Watkins, Chem Soc Rev, 9, 429, (1980)
11 H A Benisi and J H Hildenbrand, J Am Chem Soc, 71, 2703, (1949)
12 R S Mulliken, J Am Chem Soc, 72, 600, (1950)
13 R S Mulliken, J Am Chem Soc, 74, 811, (1952)
M R S Mulliken and W B Person, Molecular Complexes - a lecture and reprint volume, J Wiley and Sons Inc, London, (1969)
15 F Gutman and L E Lyons, Organic Semiconductors Part A, R E KreigerPublishing Company, Malabar, Florida, (1981)
16 L R Melby, R J Harder, W R Hertler, W Mahler, R E Benson and W E Mochel,J Am Chem Soc, 84, 3374, (1962)
17 M R Bryce and L C Murphy, Nature, 309, 119, (1984)
18 T J Kistenmacher, T E Phillips and D O Cowan, Acta-Cryst, B30, 763, (1973)
19 R E Pierels, Quantum Theory of Solids, Oxford University Press, (1955)
20 F Denoyer, R Comes, A F Garito and A J Heeger, Phys Rev Lett, 35, 445, (1975)
30
21 R L Greene, J J Mayerle, R Schumaker, G Castro, P M Chaikin, E Etemad and S J Laplaca, Solid State Commun, 20, 943, (1976)
22 Y Ueno, A Nakayama and M Okawara, J Chem Soc Chem Commun, 74, (1978)
23 T J Kistenmacher, T J Emge, P Shu and D O Cowan, Acta-Cryst, B35, 772, (1979)
24 F Wudl and E Aharon-Shalom, J Am Chem Soc, 104, 1154, (1982)
25 R C Wheland and E L Martin, J Org Chem, 40, 3101, (1975)
26 R C Wheland and J L Guison, J Am Chem Soc, 98, 3916, (1976)
27 R C Wheland, J Am Chem Soc, 98, 3926, (1976)
28 C S Jacobsen, K Mortensen, J R Anderson, and K Bechgaard, Phys Rev, B18, 905,(1978)
29 A Andrieux, C Doroure, D Jerome and K Bechgaard, J Phys Paris Lett, 40, L381,(1979)
30 K Bechgaard, C S Jacobsen, K Mortensen, H J Pedersen, N Thorup, Solid StateCommun, 38, 1119, (1980)
31 J Bardeen, L N Cooper, and J R Schieffer, Phys Rev, JB8, 1175, (1957)
32 K Bechgaard and D Jerome, Sci Am, 247, 50, (1982)
33 K Bechgaard, K Cameiro, F B Rasmussen, M Olsen, G Rindorf, C S Jacobsen, H J Pedersen and J C Scott, J Am Chem Soc, 103, 2440, (1981)
34 N Thorup, G Rindorf, H Soling and K Bechgaard, Acta Cryst, B37, 1236, (1981)
35 R J Cava, B Batlegg, R V Van Dover, J J Krajewski , J V Warszczak, R MFlemming, W F Peck Jr, L W Rupp Jr, P Marsh, A C W P James and L F Schneemeyer, Nature, 345, 602, (1990)
36 G J Ashwell, J G Allen, E P Goodings and I W Nowell, Phys Stat Sol, A, 82, 301,(1984)
37 G J Ashwell, Pr Nauk Fust Chem Org Fiz (Dr Hab Thesis) - Politech Wroclaw, 31_,3,(1986)
38 A Barraud, J Richard, M Vandevyver, K Hokzer, J Phys D, 1 9 , 2421, (1986)
39 A Barraud, L Henrion, G Derost and A Ruaudel-Teixier, Sensors and Actuators,14, 251, (1988)
31
40 B P Bespalov and V V Titov, Russ Chem Rev, 44, 1091, (1975)
41 W R Hertler, H D Hartzler, D S Acker and R E Benson, J Am Chem Soc, 84,3387, (1962)
42 R M Metzger, N E Heimer and G J Ashwell, Mol Cryst Liq Cryst, P07, 133, (1984)
43 S Akhtar, J Tanaka, R M Metzger and G J Ashwell, Mol Cryst Liq Cryst, 139, 353,(1986)
44 J S Miller and J C Calabrese, J Chem Soc Chem Commun, 63, (1988)
45 J Zyss, J Mol Electron, 25, (1985)
46 J Zyss, J F Nicoud and M Coquillay, J Chem Phys, 8^, 4160, (1984)
47 I Ledoux, D Joise, P Frenaux, J P Piel, G Post, J Zyss, T McLean, R A Hann, PF Gordon and S Allen, Thin Solid Films, 160, 217, (1988)
48 J Zyss, G J Bertier, J Chem Phys, 77, 3635, (1982)
49 K Jain, J I Cowley, G H Hewig, Y Y Cheng and R J Tweig, Optics and LaserTechnology, 297, (1981)
50 J Zyss, D Chemla and J J Nicoud, J Chem Phys, 74, 4800, (1981)
51 R Tweig, A Azema, K Jain and Y Y Cheng, Chem Phys I^ett, 92, 208, (1982)
52 G F Lipscombe, A F Garito and R S Narang, J Chem Phys, 75, 1509, (1981)
53 J L Oudar and R J Hierle, J Appl Phys, 48, 2699, (1977)
54 R J Tweig and K Jain, ACS Symp Ser, 233, 57, (1983)
55 A Aviram and M A Ratner, Bull Am Phys Soc, Jj}, 341, (1974)
56 A Aviram and M A Ratner, Chem Phys Lett, 29, 277, (1974)
57 C A Panetta, J Baghdedhi and R M Metzger, Mol Cryst Liq Cryst, 107, 103, (1983)
58 R M Metzger and C A Panetta, Synth Met, 28, 807, (1989)
59 G J Ashwell, Thin Solid Films, 186, 155, (1990)
60 G J Ashwell, E J C Dawnay and A P Kuczynski, JCS Chem Commun, 1358,(1990)
61 G J Ashwell, E J C Dawnay, A P Kuczynski, M Szablewski, I M Sandy,M R Bryce, A M Grainger and M Hasan, JCS Faraday Trans, 86, 1117, (1990)
62 G J Ashwell, J R Sambles, A S Martin, W G Parker and M Szablewski, JCS Chem Commun, 1374, (1990)
32
63 N A Bell, R A Broughton, J S Brooks, T A Jones, S C Thorpe and G J Ashwell, JCS Chem Commun, 325, (1990)
64 Synthetic Metals, Vols 41-43, (1991)Synthetic Metals, 54, (1993)
33
CHAPTER 2: SYNTHESIS AND CHARACTERISATION OF
R(4)Q3CNQ/R(4)P3CNQ*
2.1
2 . 1.1
2 . 1.2
2.1.3
2.1.3.1
2.1.3.2
Experimental
Reagents
All the reagents used were commercially available, ?nd were not purified prior to
use. The N-alkyl quinolinium and N-alkyl picolinium donors were all synthesised
from commercially available reagents.
Microanalysis
Microanalytical data were obtained from the microanalysis service of the City
University, London.
Instrumentation
Infra-red spectroscopy
Infra-red spectra were recorded as KBr discs using a Unicam SP1000 series infra
red spectrometer.
Ultra-violet/visible (uv/vis) spectroscopy
uv/vis spectra were recorded using a Perkin-Elmer 550S series spectrophotometer,
in the range 190 to 900nm, using a 1cm path length quartz cell.
FOR THIS NOMENCLATURE SEE OVERLEAF
34
Nomenclature in donor-x-acceptor adducts of TCNQ
The following shorthand nomenclature has been adopted for the zwitterions.
R - refers to the alkyl chain
(4) - refers to substitution in the '4'-position of the picolinium ring.
P - refers to picolinium donor. (Q would refer to a quinolinium donor)
3CN - shows 3 cyano groups present.
Q - refers to TCNQ as the acceptor,
eg:
R(4) P3CNQ
Efc Ofr) P3
35
Mass spectrometry
Mass spectra were recorded using a VG305 series mass spectrometer coupled to a
Casio Erba GC and VG D52050 series data system. A direct insertion probe was
used.
1H nmr spectroscopy
1H nmr spectra were recorded on a Bruker WP 805Y nmr spectrometer. Samples
were dissolved in deuteriated dimethyl sulphoxide with tetramethylsilane as an
internal reference.
Differential Scanning Calorimetry (DSC)
DSC measurements were carried out on a Mettler TA3000 systems in a dynamic
nitrogen atmosphere. Scans were made over the range -150°C to 350°C at a rate of
5°C min”1.
Synthesis
Synthesis of the N-alkyI-4-methyl quinolinium bromides
A range of derivatives was synthesised - the methods used being the same in all
cases. The synthesis of N-eicosyl-4-methyl quinolinium bromide is described below
as representative of the series.
2 .2.2
1-Bromoeicosane (13.72g, 0.038 moles) was gently refluxed with 4-methyl
quinoline (5.43g, 0.038 moles) for 2 V i hours. The initial dark brown liquid became
darker in colour and more viscous. On cooling, a dark blue solid was obtained
which was recrystallised from hot methanol. Yield 17.5g (91%), mpt 120-122°C,
X max (MeOH) 604nm.
Equation
/ /' >\
/ - A ^V
J j — H e - - ■ > R+ H clL
Synthesis of C2QH 41 (4)Q3CNQ
An examination of the structure of the quinolinium cation showed that the methyl
proton is relatively acidic and may be removed by a suitable base. Thus, the
following synthetic method was adopted.
Acetonitrile (100 cm3) was refluxed with piperidine (3-4 drops) with continuous
stirring, for 10 minutes. N-eicosyl-4-methyl quinolinium bromide (0.5g, 9.8 x 10"4
moles) was dissolved in acetonitrile (50 cm3) and added dropwise to the refluxing
solution. After addition of the cation, TCNQ (0.2g, 9.8 x 10‘4 moles) was added
and the blue solution was refluxed for a further hour, during which time the colour
changed from the intense bright blue to green. After a further 13 hours reflux, the *
solution turned turquoise and on cooling a fine, black powder was produced. The
product was isolated by filtration under suction, and washed with toluene (1 x 10
cm3) and ether (2 x 10 cm3) to remove any unreacted TCNQ. The crude product
37
was recrystallised from hot acetonitrile, filtered and dried by suction. The product
was collected as fine, dark khaki coloured particles. Yield 0.126g (21 %); mpt 286-
290°C.
Equation
C20Hu 3r + ;N v / -CH.
C20HM “ NA© / —
C H .
H
H
NC
An alternative synthetic route involves the use of LiTCNQ and a suitable alkylated
quinolinium or picolinium species, and is that quoted by Metzger et al1 in the
original paper on Me(2)P3CNQ. This method is described below.
38
2.2.3 Synthesis of LiTCNQ2
Lithium iodide (3.3g, 0.024 moles) was added to a solution of TCNQ (0.5g, 0.024
moles) in acetonitrile (100 cm3), and was refluxed for one hour. After cooling, to
room temperature, filtration gave a purple product of crude LiTCNQ, which was
washed with toluene (1 x 10 cm3) and ether (2 x 10 cm3) to remove the unreacted
TCNQ. A dark blue powder was produced in good yield (>80% ).
2.2.4 Synthesis of Me(4)P3CNQ
N-methyl-4-picolinium iodide (lg, 4.48 x 10“3 moles) dissolved in 10 cm3 of hot
water was added slowly to a refluxing solution of LiTCNQ (0.9g, 4.48 x 10“3
moles) dissolved in 200 cm3 of MeCN. The reaction mixture was then allowed to
reflux until the solution colour changed from the characteristic green of the TCNQ
radical ion to blue. On allowing to cool to ambient temperature, green
microneedles were produced which were recrystallised from acetonitrile. Yield
O.lOlg (7.95%), mpt 308 - 310°C, *Xmax (MeCN) 644nm.
39
Equation
T Li I
A collection of microanalytical data is shown is Thble 2.1.
Z w i t t e r i o n C%
C a l c ( f o u n d )
H%
C a l c ( f o u n d )
N%
C a l c ( f o u n d )
C 1 1 H 2 3 ( 4 ) Q 3 C N Q
C 1 4 H 2 3 ( 4 ) Q 3 C N Q
C 1 5 H 3 1 ( 4 ) Q 3 C N Q
C 1 6 H 3 3 ( 4 ) Q 3 c N Q
C 2 0 H 4 1 < 4 1 Q 3 C N Q
8 0 . 9 7 ( 8 1 . 0 0 )
8 1 . 2 0 ( 8 0 . 4 6 )
8 1 . 4 6 ( 8 1 . 2 6 )
8 1 . 5 8 ( 8 0 . 5 4 )
8 1 . 9 5 ( 8 1 . 1 0 )
7 . 2 1 ( 7 . 4 5 )
7 . 9 8 ( 7 . 7 5 )
7 . 9 7 ( 8 . 1 4 )
8 . 1 4 ( 8 . 0 1 )
8 . 7 2 ( 9 . 1 1 )
1 1 . 8 0 ( 1 1 . 7 2 )
1 0 . 8 2 ( 1 0 . 1 3 )
1 0 . 5 5 ( 1 0 . 5 6 )
1 0 . 2 8 ( 1 0 . 0 8 )
9 . 3 2 ( 9 . 0 2 )
T&ble 2.1 Microanalytical data for selected zwitterionic adducts
40
2.3 Discussion of Synthetic Methods
The synthesis of R(4)Q3CNQ and R(4)P3CNQ, where R is an alkyl chain, is a
substitution reaction of TCNQ in which a cyano group is substituted by the
appropriate quinolinium/picolinium donor. The two alternative methods have
differing mechanisms which are shown in fig 2.1 and fig 2.2.
The first method involves the use of piperidine as a base to remove a proton from
the 4-methyl group of the N-alkyl-4-methyl-quinolinium ion, a reaction very similar
to the preparation of merocyanine dyes using N-alkyl picolinium salts with
piperidine again used as a base3. The alternative mechanism presumably uses the
LiTCNQ as the base, which also removes a proton from the 4-methyl group.
The removal of the proton results in the formation of a stable species which has a
resonance structure in which the lone pair of electrons is situated on the electron
deficient nitrogen. Such a methyl proton abstraction results in an extremely
reactive homo-aromatic base which is highly susceptible to electrophilic addition at
the exocyclic carbon. The use of acetonitrile as opposed to a polar protic solvent -
say, methanol for example - is preferable as methanol would have a tendency to
hydrogen bond to the carbanionic resonance structure.
The yields obtained by each reaction type were not particularly large, though using
piperidine they were markedly in excess of those using LiTCNQ. This would be
expected on the basis of the proposed mechanism, piperidine being a much stronger
base than LiTCNQ. Steric factors also increase the rate of reaction, as again would-
be expected, with a noticeable increase in reaction time occurring as the length of
the alkyl chain is increased. This effect is particularly noticeable using LiTCNQ.
The preparation of CioFi7H4(4)P3CNQ was eventually carried out using this
method, but only after heating under reflux for two months. Such a reaction is
41
Fig 2.1Proposed mechanism for the synthesis of R(4)Q3CNQ using piperidinebase
CN
NC
Y
42
T h e n
R -
CN
R - N \ /
43
O r
44
Fig 2.2 Proposed mechanism for the synthesis of R(4)Q3CNQ using LiTCNQ
45
clearly unsatisfactory if the synthesis of long chain analogues is to be carried out on
a routine basis. For these reasons, the method described in section 2.2.2 was used
in the majority of the synthetic work, even though the problems of additional side
reactions can cause the final product to be somewhat impure.
The progress of the reaction could be monitored by observing the uv/vis spectrum
of the reaction mixture (for a more detailed discussion of the uv/vis spectra of these
materials see Sec 2.4.1). Initially, the spectrum shows bands characteristic of the
TCNQ° radical anion at ca. 840nm, though as the reaction proceeds, a broad
charge transfer band characteristic of the particular zwitterionic adduct can be
observed centred around 600-750nm. An example of the change in uv/vis spectrum
of the reaction mixture during the course of the reaction for CioFi7H4(4)P3CNQ
is shown in fig 2.3. The reaction can also be monitored using thin-layer
chromatography (TLC). Using a 10:1 methylene chloride:methanol solution as the
mobile phase, pure (one-spot) zwitterionic products can be identified having an R̂ .
value of 2.2. Thus, these two methods, taken together provide adequate means of
monitoring the course of the reaction.
46
- 50
- 40
- 30
- 20
500 700 900
Wavelength (nm)
S t a r t i n g M a t e r i a l s
- 70
- 60
- 50
- 40
- 30
- 20
3c: 500 700 900
Wavelength (nm)
Absorbance
Absoroanoe
Fig 2.3 UY/VIS spectra showing the course of the reaction between LiTCN*
the C10F17H4 picolinium cation
- 70
- 60
- 50
- 40
- 30
- 20
50030C 700 900
Wavelength (nm)
- 60
- 50
- 40
- 30
- 20
500 930700
Wavelength (nm)
Absorbance
Absorbance
48
- 70
- 60
- 50
- 40
- 30
“ 20
- 1 0
500300 900700
Wavelength (nm)
- 70
- 60
- 50
- 40
- 30
- 20
- 10
300 500 900700
Wavelength (nm)
Absorbance
Absorbance
49
- 70
- 60
- 50
- 20
300 500 700 900
Wavelength (nm)
- 70
- 60
- 50
- 40
- 30
r 20
300 500 700 sooWavelength (nm)
Pure C10F,7H4 (4)P3CNQ
A r s o r b a r ic e
Acsorhance
50
2.4 Spectroscopic Studies of R(4) or R(2)P3CNQ/R(4) or R(2)Q3CNQ
2.4.1 Ultraviolet/visible spectroscopy
All pure adducts show an intense, broad transition centred around 600-750 nm in
acetonitrile solution. Two less intense transitions at ca 340 and 378nm are also
observed. Typical spectra are shown in fig 2.4 and 2.5. The low energy transition
is independent of chain length, and has been attributed to an intramolecular back
charge transfer process from the negatively charged dicyanomethanide group to the
positively charged quinolinium moiety1. This is substantiated by plots of
absorbance against concentration4 which have been shown to obey Beers Law. As
this law is dependent on there being no interaction between absorbing
chromophores, such behaviour can be taken as a good indication of an
intramolecular process. Thble 2.2 shows ^ max(nm) values for a range of
quinolinium and picolinium analogues.
As would be expected, the energy of transition is independent of chain length - the
active portion of the molecule being the same for each zwitterionic system.
However, inductive effects due to increased electron withdrawing character of the
alkyl chain, manifest themselves in changes in "X max. Consider the molecule
CioFi7H4(4)P3CNQ, where the very electronegative fluorine atoms tend to draw
electrons away from the picolinium donor. In this case the "donating" strength of
the molecule is diminished slightly. Thus, the intramolecular back charge transfer
process from acceptor to donor requires less energy and the absorption maximum is
shifted to longer wavelength. This effect is shown in Table 2.3.
51
2de
1
a
Fig 2.4 UV/VIS spectra of C,Hlf(4)Q3CNQ in MeCN
52
saw
- 7 0
- 6 0
- 5 0
- 4 0
- 3 0
- 20
- 10
3 0 0 5 0 0 7 0 0 9 0 0
W a v e l e n g t h ( n m )
F i g 2 . 4 ( C o n t d . . ) U V / V i s s p e c t r a o l C „ H 23 ( 4 ) Q 3 C N Q i n M e C N
• A b s o r b a n c e
53
h 7 0
b 6 0
b 5 0
b 4 0
b 3 0
b 20
3 0 0 5 0 0 7 0 0 9 0 0W a v e l e n g t h ( n m )
F i g 2 . 4 ( C o n t c L ) U V / V i s s p e c t r a o f C I 3 H , 7 ( 4 ) Q 3 C N Q i n M e C N
A b s o r b a n c e
54
zm
Fig 2.4 (Contd..) UV/Vis spectra of CnH „(4)Q3CNQ in MeCN
55
9 0
8 0
7 0
5 0
4 0
3 0
3 0 0 7 0 05 0 0 9 0 0
W a v e l e n g t h ( n m )
F i g 2 . 5 U V / V i s s p e c t r a o f M e ( 2 ) P 3 C N Q i n M e C N
A b s o r b a n c e
56
8 0
7 0
6 0
5 0
4 0
3 0
20
— i—
5 0 0I
7 0 0|
9 0 03 0 0
W a v e l e n g t h ( n m )
F i g 2 . 5 ( C o n t d . . ) U V / V i s s p e c t r a o f C a H l7 ( 4 ) P 3 C N Q i n M e C N
A b s o r b a n c e
57
- 7 0
- 6 0
- 5 0
- 4 0
- 3 0
- 20
- 10
3 0 0 5 0 0 9 0 07 0 0
W a v e l e n g t h ( n m )
F i g 2 . 5 ( C o n t d . . ) U V / V i s s p e c t r a o f M e ( 4 ) P 3 C N Q i n M e C N
A b s o r b a n c e
58
•VCD:'a>
0c
Big 2.5 (Contd..) UV/Vis spectra of C10F17H,(4)P3CNQ in MeCN
59
A D D U C T m a x ( M e C N / n m )
M e ( 2 ) P 3 C N Q ' 5 9 9
M e ( 4 ) P 3 C N Q 6 4 4
E t ( 4 ) P 3 C N Q 6 4 9
M e ( 2 ) Q 3 C N Q 6 9 6
M e ( 4 ) Q 3 C N Q 6 9 8
C g H 1 7 ( 4 ) P 3 C N Q 6 4 9
C 1 0 F 1 7 H 4 ( 4 ) P 3 C N Q 6 7 2
C g H 1 7 ( 4 ) Q 3 C N Q 7 0 8
C 1 0 H 2 1 ( 4 ) Q 3 C N Q 7 1 0
C 1 1 H 2 3 ( 4 ) Q 3 C N Q 7 0 0
C 1 5 H 3 7 ( 4 ) Q 3 C N Q 7 1 0
C 2 0 H 4 1 ( 4 ) Q 3 C N Q7 1 0
Table 2.2 Selected spectroscopic (UV/vis) data for zwitterionic adducts of TCNQ
60
Z w i t t e r i o n " A m a x ( n m ) M e C N
E t ( 4 ) P 3 C N Q 6 4 9
C 8 H 1 7 ( 4 ) p 3 ° N Q 6 4 9
C 1 0 F 1 7 H 4 1 4 ) P 3 C N Q 6 7 2
Table 2.3 Variation in A max with increasing electron with drawing character of the alkyl chain
2.4.1.1 Solvatochromism in R(4)/R(2)P3CNQ and R(4)/R(2)Q3CNQ5
The zwitterionic adducts exhibit solvatochromism, that is, the values of A max are
solvent dependent. The phenomenon of solvatochromism in related merocyanine
dyes has recently been reviewed6. Such solute-solvent interactions can be due to a
number of factors:
1) Dipole-dipole interactions where the solute and solvent are polar.
2) Solute dipole-solvent induced dipole interactions where the solute is polar and
the solvent is not.
3) Solvent dipole-solute induced dipole interactions where the solvent is polar
and the solute is not.
4) Dispersion interactions.
5) Specific interactions such as hydrogen bonding and charge transfer complex
formation with the solvent.
6) Solvent cage compression effects.
61
Factors 1-4 are general electrostatic phenomena; however, where 5 and 6 are
present, these will predominate. Much work has been done in the past711 - both
experimental and theoretical - attempting to explain such interactions. It is not
within the remit of this work to discuss this in any great detail. However, the
following points are relevant to the following short discussion.
It is predicted that6 non-polar solutes should undergo a bathochromic (red) shift as
the solvent polarity increases. For polar solutes the situation is more compicated in
that solvatochromism in such species is dependent on the solute dipole moment in
the ground state and its change on excitation.
Expanding on this in more detail, then it has been shown that an increase in dipole
moment on excitation should lead to a bathochromic (red) shift. Neutral
chromophores may exhibit this behaviour. A decrease in dipole moment (for
example charged chromophores) on excitation should lead to hypsochromic (blue)
shifts being observed with increasing solvent polarity. The variation in ^ max for
various adducts in a variety of solvents is shown in figs 2.4 and 2.5.
62
Wave
leng
th
Maxi
ma
(nm)
740
7 2 0
70 0
6 0 0
6 0 0
0 4 0
07.011
IPA
MeOH
2 0 30D i e l e c t r i c C o n s t a n t
Figure 2 A Variation ofAmax with dielectric constant o f solventfor M c(4)Q 3C N Q
CII3CN
J4 0
63
wave
leng
tn
.Maxzma
000 f-
700
6 0 0
x n / C N
- IPA
Cl 13) 2C0
LOU
DMP
MoOII
10 20 30 <10D i e l e c t r i c C o n s t a n t
Figure 2.5 Variation o a max with dielectric constant or solvent fo r C 10F17IIJ 4 )P 3 C N Q
64
As can be seen, the theoretical predictions are bom out by the results obtained. The
polar adducts, which undergo a reduction in dipole moment on charge transfer,
show clear hypsochromic shifts with increasing solvent polarity. Notice also that
the effects of hydrogen bonding appear apparent in both figures - the protic solvents
following an independent trend. The max value of CioFi7H4(4)P3CNQ in
cyanobenzene appears not to follow either trend, though this may be due to some
weak charge transfer interaction.
2.4.2 Infra red spectroscopic studies
As expected, all the adducts give similar infra-red spectra. Typical spectra are
shown in figs 2.6 and 2.7 with all the relevant frequency assignments in Table 2.4. .
The important peaks are the two - C = N stretching bands at 2137 cm-1 and
2177 cm'1. Neutral nitriles normally cover the range 2260 cm "1 to 2200 cm "1.
The presence of the two cyano stretching frequencies are characteristic of the two
distinct cyano environments12' 14. The band at 2177 cm "1 represents the neutral
environment whilst the band at 2137 cm"1 is characteristic of the dicyanomethanide
3-carbon unit over which the negative charge is delocalised. Thus this
delocalisation of charge reduces the strength of the bond and the stretching
frequency is reduced correspondingly.
65
Figure 2.6
Infra red
spectra for
CnH
23(4)Q3C
NQ
TRANSMITTANCE (%ioo o oo __
O p e r r—i F—■
C 5
CJloo
66
WA
VELEN
GTH
(M
ICR
ON
S)
Figure 2.7
Infra red
spectra for
CisH
3i(4)Q3C
NQ
TRANSMITTANCE { % )
B I
O
—
CDoo _
"0oo
;i __ i i
- - * ^, | : : CZli l i • , i i
i ’ :1 >
1 > 1 *1 • i , i i • i !
1 • I . I i 1 ‘ i \ - i l ii i i « i i i . i • i i i i M
too o CDo oo
67
WA
VELEN
GTH
(M
ICR
ON
S)
A b s o r p t i o n B a n d A s s i g n m e n t
2 9 2 0 ( s ) - C - H s t r e t c h
2 2 0 0 ( v s ) - C = N s t r e t c h ( n e u t r a l )
2 1 5 0 ( v s ) - C = N s t r e t c h ( i o n i c )
1 6 0 0 ( S ) - C = C s t r e t c h ( a r o m a t i c )
1 6 5 0 ( S ) - C = C s t r e t c h ( n e u t r a l )
1 5 0 0 ( S ) - C = C s t r e t c h ( a r o m a t i c )
Table 2.4 Infra-red spectral bands for an R(4)Q3CNQ adduct (s) strong; (vs) very strong
68
2.4.3 Mass Spectroscopy Studies
Mass spectroscopic data for a range of quinolinium and picolinium adducts,
interpreted. Several interesting fragmentation pathways of some diagnostic
importance have been identified.
Examples of the spectra obtained are shown in figs 2.8 - 2.14.
The basis of the technique is the production of ions from neutral species.
Bombardment of the molecule with high energy electrons leads to the formation of a
positively charged molecular ion. This molecular ion may then break into smaller
fragmentation ions. The processes can be summarised as shown below:
1) M + e ------------------- > M-+ + 2e
ie loss of an electron leads to a radical cation.
substituted in the '2 ' and '4 ' position of the donor ring have been obtained and
6
69
Figure 2.8
Mass spcclra
of Mc(2)P3C
NO
3Z129v 99 99 991
t-
'■ r -
£ - -
w .ji. =
£
L
VSM
i
L
70
IB
Figure 2.9
Mass
spectra of I*t(2)P
3CN
O
[79 9c 9* 99 3 9 3 9 ;
#0<so
71
Figure 2.10
Mass spectra
of r*t(4)I>3CN
Q
9w12
99 99 9 9 :
72
LS
?
figure 2.11
Mass
spectra of C
iol I23(4)Q3C
NQ
9 *
; 239 98 9 3 :
73
ssr
Figure 2.12
Mass
spectra ofC
i5H3i(4)Q
3CN
Q
74
Figure 2.13
Mass
spectra of C
nl I23(4)03CN
0
* c:T* T » c:
t£-
i t
r*tr
L
rr i
©
■L
75
Figure 2.14
Mass
spectra of C
20I Ui(4)Q
3CN
Q
76
5U
Pathway A leads to a new radical cation and a neutral molecule as shown, and B to
a cation and a radical.
Bearing these principles in mind, the spectra can be analysed. For ease of
interpretation, the molecular ion is best regarded as involving the quinoid structure
as shown in figure below.
RMM = 329 + 'R '
Consider the spectra shown in figures 2.8 and 2.9. These quite clearly show the
loss of the neutral molecule HCN where the a-position of the
quinolinium/picolinium donor is available. Fig 2.10 shows the spectra of
Et(4)P3CNQ and such substitution in the
^ -position does not lead to loss of HCN. The following mechanism shows why
this is the case:
77
RM M 284
RM M 2 5 1
4-Hcn RM M 27
This mechanism, resulting as it does, in the production of a five-membered ring,
can only occur when substitution occurs in the a-position and not the X -position.
When applicable, such a fragmentation pathway is favourable due to the relative
stability of the five-membered transition state and the resulting formation of product
ion and neutral HCN.
In both picolinium and quinolinium, a / X -substituted adducts, where the alkyl chain
is of 2-carbon atoms or more, then loss of an alkene, as a neutral molecule, occurs.
Again, the proposed mechanism involves a cyclic transition state - as shown over
page, and these fragmentations are identifiable in the mass spectra of
Ci 5H3 i ( 4)Q3CNQ 2-12)> Cn H31(4)Q3CNQ (fig 2.13) and C20H41(4)Q3CNQ
(fig 2.14).
All R(4)Q3CNQ/R(2)Q3CNQ adducts show a strong peak at RMM 255 and
R(4)P3CNQ/R(2)P3CNQ at RMM 205. It is proposed that these fragments'
correspond to the structures shown in Figure 2.15.
79
For example, from figure 2.11
V i / *
H C ^
H - , CL I
H
R
c J 0 H 2 1 ( ^ ) q 3 C N G L
R MM 4 6 0
N\
H
H2C^C _
/ H
H
N C t N
RMM 320
+ C8H 16CH = CH.
RMM 140
80
RMM 255
RMM 205
Figure 2.15 Proposed fragmentation structures
81
The mechanism by which these structures are produced is outlined over the page
and involves several fragmentations. Notice that the double bond is shown
promoting reactivity - a mechanism which would appear unrealistic in the presence
of the generally more reactive cyano groups. However, the mass spectroscopy
technique involves such a high energy regime that such reactions cannot be ruled
out.
Step '2 ' involves the loss of a carbene, - a highly reactive intermediate and thus a
favoured process. Step ' 1' involves the loss of the corresponding alkyl chain and is
common to all the spectra obtained.
To conclude, distinct differences in the fragmentation patterns are apparent between
a and ^ - substituted adducts which can readily be explained by facile mechanisms.
82
1)
Molecular ion RMM 284
2-)
C u H9N2 RMM 205
83
2.5 Miscellaneous Studies
Thermal analysis of various charge-transfer complexes of TCNQ, using differential
scanning calorimetry (DSC), have shown that certain interesting features can be
explained by the presence of solvent trapped in the crystal lattice.15 For the
production of good quality LB films for molecular electronic applications, this
would be undesirable.16 Thus, materials were studied using DSC, and no phase
changes corresponding to the presence of trapped solvent were observed.
84
2.6 REFERENCES
1 R M Metzger, N E Heimer and G J Ashwell; Mol Cryst Liq Cryst, 107, 133, (1984)
2 D S Acker and W R Hertler; J Am Chem Soc, M , 3370, (1962)
3 I R Girling, Personal Communication
4 M Szablewski, MPhil thesis, Cranfield Institute of Technology, (1989)
5 K M C Davis in "Molecular Association" vol 1, Ed.R Foster, Academic Press (1975)
W Liptay, Angew Chem Int Ed Engl, 8, 177, (1969)
6 E Buncel and S Rajagopal, Acc Chem Res, 23, 226, (1990)
7 C Reichardt, Solvent Effects in Chemistry, 2nd Ed, Verlag Chemie, Weinheim,(1988)
8 H W Gibson, Tetrahedron, 77, 6789, (1977)
9 E M Kosower, J Am Chem Soc, 80, 3253, (1958)
10 L G S Brooker, A C Craig, D W Heseltine, P W Jenkins and L L Lincoln, J AM Chem Soc, 87, 2443, (1965)
11 A Botrel, A L Beuze and P Jacques, J Chem Soc Faraday Trans 2, 80, 1235, (1984)
12 J S Chappell, A N Bloch, W A Bryden, M Maxfield, T O Poehler and D O Cowan, J Am Chem Soc, 103, 2442, (1981)
13 A Terzis, E I Kamitsos, V Pscharis, J S Zambounis, J Swiatek and J Papavassilou, Synth Met, 19, 481, (1987)
14 Basic Infrared Spectroscopy, J H van der Maas, 2nd Ed, Heyden & Son Ltd, (1972)
15 G J Ashwell, I M Sandy, A Chyla and G H Cross, Synth Met, 19, 463, (1987)
16 R A Hann in 'Langmuir-Blodgett Films' Ed G G Roberts, Academic Press, (1990)
85
CHAPTER 3: LANGMUIR-BLODGETT FILMS’-2
3.1 Historical Introduction
The phenomenon of oil on water has been known for many years. The Babylonian
people, thousands of years ago, apparently observed the spreading of oil on water
as a form of divinity.3 Of perhaps a more practical nature, the Japanese art form
s u m o -n a g a s h i involves the passing of paper through a spread suspension of carbon
and protein molecules on a water subphase, to produce a patchwork design of dark
and light areas.4 The calming influence of "oil on troubled water" has also been
known to fishermen throughout the ages.
However, the first scientific experiments on monolayers are accredited to the
famous American Statesman and scientist, Benjamin Franklin. He reported to the
Royal Society way back in 1 1 1 A5 that a teaspoon of oil dropped onto the pond on
Clapham Common extended to an area of approximately half an acre. In common
with the observations of mariners, he also reported on its calming influence. It was
Lord Rayleigh,6 some years later, who deduced that such a volume of oil, spread
over such an area would produce a layer one molecule thick.
Concurrent with the work of Lord Rayleigh, a German schoolgirl, Agnes Pockels,7
carried out a series of experiments determining molecular sizes. Out of this work,
developed the apparatus which was the basis of what is now known as the Langmuir
trough - although the actual apparatus she used amounted to little more than the
kitchen sink.
The true significance of this seminal work by Rayleigh and Pockels went unnoticed
for some years until a scientist working at the General Electric laboratories in
Schenectady, New York, Irving Langmuir, fully developed the theory of
8 6
monolayers, work for which he was awarded the Nobel Prize. In an early
communication,8 he described the design of his film-balance, or trough, and showed
how it could be used to deduce the molecular size and orientation of monolayers at
the air-water interface.
The first description in detail of the sequential build up of monolayers onto a solid
substrate was described by Blodgett9 and such assemblies are now called Langmuir-
Blodgett (LB) films. With the advent of the Second World War, the research
interest into LB films ceased, and it was not really until the I960's that the topic
created renewed study. This was due to the work of a German group led by Hans
Kahn who published a series of papers10 on the effect of electron transfer in
supermolecular assemblies. Since then, work in this field has continued apace, with
five international conferences' - in England, America, West Germany, Japan and
France having taken place since 1983, and the sixth, in Canada, due later this year.
The microelectronics industry is currently very interested in thin organic films.
Currently, the industry relies on inorganic materials for the majority of its needs,
and the number of suitable materials is limited. There is a wealth of organic
materials already known and countless others waiting to be synthesised. The ability
of organic chemists to "tailor-make" molecules fitting a specific molecular
requirement opens exciting possibilities - could these molecules be suitably
fabricated? This, coupled to the obvious need for very much thinner, defect-free
films of a precise order and thickness, is the attraction of LB film research. What,
therefore, are the characteristics of those molecules suitable for study as potential
LB film materials?
87
Molecular Requirements of LB film forming materials
Classically, much of the work done on LB films involved the use of long chain fatty
acids - shown schematically in fig 3.1 - of which stearic acid was one of the most
widely studied.
In stearic acid there are seventeen carbon atoms constituting a hydrophobic (water-
hating) hydrocarbon "tail", and a hydrophilic (water-loving) carboxyl "head" group.
Such molecules are said to be amphipathic and it is this amphipathic balance
between hydrophobic and hydrophilic moieties that is significant.
Cr7 H j j CCCH
H
c
H C H
H
Figure 3.1 Schematic diagram of a long chain fatty acid
Other materials of this type have been widely studied including, for example,
arachidic acid (C19H39COOH) and behenic acid (C21H43COOH). Thus from this
early work, the key requirement was found to be a hydrophilic and hydrophobic
portion within each molecule. Some more specific structural types are now
discussed.
The introduction of double bonds into the hydrocarbon chain has been achieved.
For example,co -tricosenoic acid (fig 3.2) has been extensively studied. The
presence of the terminal double bond in this material greatly reduces the disruption
of the close-packed hydrocarbon chains, which is commonly observed when the
double bond is situated mid-chain.
LB films of this material have been extensively studied due to their polymerisation
when exposed to an electron beam.11
CH j * CH - (CHjIto- c o c h
V
H C Hitiii
H C H
<AoiH
Figure 3.2 c*-tricosenoic acid
As well as being easily polymerised, such films exhibit excellent thermal and
mechanical stability. The very similar alkyne derivative (fig 3.3) also readily
undergoes polymerisation.
89
Figure 3.3 An alkyne derivative
It is a relatively straight forward task for the organic chemist to attach such long
hydrophobic tails to many of the suitable hydrophilic head groups and then spread
the material as a monolayer. Indeed, the number of film forming materials now
known is vast and increasing steadily. These include variations upon the
hydrophilic portions, including, for example, cyano13 and sulphur14 containing
materials. The ability of preformed polymers to produce stable films is also well
documented, both in early work15 and more recent studies.16 It has also been
shown that many of the organic charge-transfer complexes described in Chapter 1
can produce LB films, in particular TCNQ-pyridinium salts17 and TCNQ-TTF,18
the first organic metal.
The attachments of long hydrophobic groups severely restricts any practical
applications. Two factors may explain this: firstly, the presence of such large
hydrophobic groups can introduce a large degree of instability and, secondly, the
more interesting properties of the organic molecule can be severely diluted by the
long aliphatic chain. Recently, therefore, more robust molecules containing much
less aliphatic character have been fabricated as LB films. These include the
phthalocyanines (fig 3.4), first studied at Durham19 and porphyrins20 (fig 3.5).
Figure 3.4 Basic structure of a Phthalocyanine molecule
MJ
P2
Figure 3.5 A Porphyrin derivative
91
Study of such species have shown that the films produced have a generally imperfect
structure when compared to the long chain fatty acids, though this is perhaps the
price to be paid as a consequence of greater mechanical and thermal stability.
Recently the ad hoc nature of the search for new materials has been somewhat
superseded by more coordinated work on structure property relationships. Perhaps
the first schematic study of this type was carried out on anthracene derivatives by
scientists at Durham and ICI.21 It was shown that systematic alteration of the
hydrophobic and hydrophilic portions could enable molecules of the structure shown
in fig 3.6 to be fabricated as LB films. Here R' is a carboxylate group and R can
be as short as four carbon atoms.
R
H
H
R*
Figure 3.6 A C-4 anthracene derivative suitable for fabrication as an LB Film
Thus, one conclusion from this work was that the size of the aliphatic chains could
be reduced if a polycyclic aromatic system was incorporated.
92
To conclude, therefore, the number of suitable materials is now extensive and work
is now following a more systematic path as regards the often subtle external changes
needed to effect gross changes in film behaviour.
3.3 Isotherm Measurement
Amphipathic molecules, dissolved in an appropriate solvent, may be spread upon a
suitable subphase. The solvent then evaporates and the remaining molecules are
compressed by means of a barrier. The resultant plot of surface pressure against
area per molecule (an isotherm) obtained during the compression of a long chain
fatty acid is shown in fig 3.7.
uoLlJ
OLUcn
c cc l .
u j
25 30 35 40A R E A / M O L E C U L E (A/Mri )
Figure 3.7 Surface Pressure vs area per molecule
Characteristics for a long chain fatty acid
93
The floating monolayer is a two-dimensional system and can be described as a two-
dimensional gas obeying the equation
PA = kT(Eq6)
where 'P* is the surface pressure, 'A ' the effective area per molecule, 'k ' the
Boltzmann Constant and 'T ' the thermodynamic temperature. On compressing the
monolayer the surface pressure increases and three distinct phase changes are
observed, from gaseous through liquid to a condensed solid phase as identified in
fig 3.7. However, it must be emphasised at this stage that many materials
producing stable, good quality LB films do not produce the classic, three-phase
isotherms obtained for fatty acids. For example, Barraud et al22 have shown that
the plateau observed in the isotherms studied during their extensive analysis of
TCNQ-pyridinium salts can be explained by the formation of dimers on the
subphase. The studies by Lewis et al14 on sulphur containing molecules have also
shown non classical behaviour, in this case, the anomalies are explained by gross
reorientation on the water subphase of the amphiphile during compression. Indeed,
the complexity of most LB film forming materials compared to the fatty acids would
lead one intuitively to such conclusions.
Extrapolation of the linear part of fig 3.7 to point ‘E ', as shown, characterises the
area at zero pressure and is taken to be the cross-sectional area of the molecule on
the water subphase. From a consideration of this limiting area per molecule and the
actual form of the isotherm, important data on the behaviour and suitability of a
particular molecule can be obtained. As such, an important pre-requisite to the
deposition of LB films is a full characterisation of the material on the subphase.
94
The quality of any isotherm obtained is crucially dependent on the quality of the
subphase.23 The purest quality water available is generally used. Most commercial
water purification systems involve distillation processes, ion-exchange stages and
reverse osmosis, though the exact requirements are dependent on many factors,
including local water quality. (The exact system used in this study is discussed in
section 4.6).
Recently there has been interest in the use of alternative subphases, in particular the
use of glycerol to study, amongst other compounds, TCNQ based molecules24. In
this case the use of glycerol reduces monolayer solubility and enables the use of
solvents generally prohibited due to their solubility in water.
The presence of divalent metal ions can, in many case, aid the cohesion of
conventional fatty acid and other charged monolayers. The resultant structure of
such layers, and the self-explanatory reason for this, is shown in fig 3.8.
H- C- H
0 ”
Figure 3.8 Molecular arrangement of fatty acid salts at the air-water interface
95
Such materials with ionisable head groups are also dependent upon the pH of the
subphase - the presence of a suitable acid preventing the hydrolysis of the material.
The combined effects of adding divalent metal ions and pH are shown in fig 3.9.
The quality of the monolayer is critical and it is this quality that eventually governs
the properties of the film on a solid substrate. A major source of problems are the
formation of molecular domains. These lead to the formation of imperfect films
which, in the case of electrical applications, may not be able to withstand large
electrical fields. The presence of domains can be substantially reduced if slow
evaporating solvents are used to spread the material, for example mesitylene and
long chain hydrocarbons.
The experimental consideration and precautions taken for producing reproducible
isotherms will be discussed in greater length in Chapter 5.
96
3.4 Deposition and Mono/Multilayer Assembly
The fabrication of an LB film involves the lifting of a suitable substrate vertically
through a compressed monolayer, maintained at a predetermined surface pressure.
Repeating this process a successive number of times produces a multilayered
structure.
This vertical dipping method produces three distinct deposition modes - X, Y and Z
types. The structure of these is shown in fig 3.10. As is clear, on each deposition
cycle material may, or may not, be transferred to the substrate. The transference of
material to the substrate manifests itself as a reduction in the area occupied by the
film - the surface pressure necessarily remaining constant. Thus, the transfer ratio
can be defined as:
decrease in area of subphase covered by film
T area of substrate covered by the monolayer
The value of X j gives a good indication as to whether transference is occurring
correctly - ideally X j should be unity or zero. Values differing substantially may
imply rearrangement of molecules or incomplete pick-up.
With reference to fig 3.10 then:
a) X j = 0 and XJ = 1, X-type deposition
b) XJ = 1 and XJ = 1, Y-type deposition
c) X j = 1 and Tj = 0, Z-type deposition
Notice that X- and Z-type depositions are associated with a single layer unit cell in
a direction normal to the film surface, and that the Y-type is associated with a
bilayer unit cell.
98
Figure 3.10 Vertical Dipping Technique
a)
b)
c)
a) X-type deposition"
b) Y-type deposition"
c) Z-type deposition"
Y-type deposition is by far the most common mode, though due to the non-
centrosymmetric nature of X-type and Z-type there is considerable interest in these
structures. However, as yet there is no clear understanding as to why some
molecules deposit in this way.
Recently, workers have used a horizontal lifting method in order to transfer material
in the gaseous and liquid monolayer states.25 As its name implies, the substrate is
held horizontally and lowered onto the monolayer from above, as shown in fig
3.11.
1 1U U O u (J u j J c3
Figure 3.11 The Horizontal Lifting Langmuir-BIodgett Technique (HLLB)
In conclusion, therefore, the LB method for production of organised assemblies,
free from vacuum processes and heating, provides a suitable mechanism for the
production of very thin organic films of a defined structure and thickness.
100
3.5 Characterisation of LB Films
The methods available for the characterisation of LB films are many and varied,
ranging from the very simple to the very complex. This short discussion shall start
by describing the methods available for elucidating the structure of the film on the
subphase and what this can tell us, proceeding then to the LB film itself - its
morphology, molecular orientation, etc.
As has probably been made clear already, a great deal of information about the film
forming potential of a particular molecule can be obtained simply from basic PA
isotherms. These can show whether a particular molecule forms a film at the air-
liquid interface at all, and if it does, whether the film is liquid, liquid-condensed, or
condensed. It may also show the presence of any distinct phase changes undergone
on compression. The area per molecule value can give a good indication of the
molecular orientation, particularly if the value obtained is rather different from the
value predicted on the basis of the known molecular structure. Whilst noTin
isolation offering conclusive proof, gross changes from expected PA behaviour - for
example the observation of plateaus when P is essentially constant for sizable
decreases in A - has led researchers to suspect that there is gross molecular
reorientation occurring and some association between molecules, as discussed
earlier. However, such observations could also be indicative of surface or
compound contamination, and thus a certain amount of caution should be exercised.
However, for the full potential of LB films to be realised then it is the structure of
the films themselves that need to be of a particular integrity, and it is the techniques •
used to investigate these that much of the research effort has been targeted.
Fortunately, many of the techniques available to solid state physicists, and
developed over a number of years, are applicable to LB film characterisation..pa
Preliminary measurements on the thickness of the films are useful as these can tell
101
whether the films have been transferred with some integrity. For example, it can be
shown that the molecule is lying purely perpendicular to the substrate or whether it
is tilted to some degree, depending on the interlayer spacing determined. Basic
optical techniques, for example Normarski microscopy26 can be used in this respect.
This technique is sensitive to the change in thickness either side of a step which is
manifested as a phase difference in the emergent wave. Thus, as each layer will
represent a "step", the thickness can be determined. Other techniques involve
ellipsometry,27 plasmon surface polariton field measurements,28 though by far the
most sensitive technique, giving the most accurate and detailed information comes
from low-angle X-ray reflection.29 This can determine the thickness down to a
single layer with an accuracy better than 1A As well as giving details on the basic
film structure, its technique can also be used to probe orientational changes in films
when undergoing a chemical reaction. It has been particularly useful in elucidating
the chemical change involved when a charge transfer salt of TCNQ, deposited as an
LB film, is exposed to iodine vapour. Such salts are found to be insulating films
before exposure to the iodine. On exposure, the films acquire a conductivity and
shifts in X-ray peaks are indicative of changes in film thickness and an overall
ordering of the film due to localisation of the iodine.
Other techniques applicable to the determination of intermolecular thickness are the
variety of electron diffraction methods available. The high energy techniques are
more commonly used either through the film - transmission electron diffraction
(TED),30 or orientated at a particular angle - reflection high-energy electron
diffraction (RHEED).31 These techniques are particularly useful as only the
outermost layers of the films are studied and thus the technique is relatively non
destructive. Both techniques are sensitive even to a single layer. This is not to say
low energy electron techniques cannot be used - indeed a single layer fatty acid film
has been investigated using low energy electron diffraction (LEED).32 As the
wavelength of the electrons in this technique are similar to those of X-rays then
102
similar results could be expected to be obtained.
Study of the films by any of the techniques mentioned above will probably have
revealed that the film is not perfect and that certain inhomogeneities and defects
exist. So techniques are available for the examination and study of such film
defects. Conventional microscopy is of some use in this area though overall its
applicability is limited. Defects do not scatter light and thus observation of them
with an optical microscope is apparent only by phase differences related to thickness
differences between "pure" film and defect. If crystallites are present, then
polarised light can be used which can deduce the birefringent character of such
species.
For direct observation of the films, however, then scanning electron microscopy
(SEM) is applicable.33 This has even been used to observe a single layer. Though
this may appear strange, because the penetrating depth of such electrons would be
deep into the substrate and thus the film should be masked by this, there are
differences in the secondary electrons emitted when the surface is covered with a
monolayer. However, this technique has drawbacks, and a certain amount of
caution should be used, particularly as to whether the film can withstand the
treatment process and the high energy beams used for film visualisation.
Transmission electron microscopy (TEM)34 has also been applied to this area and
one great advantage of this technique is that it enables the texture of the film and the
crystallographic pattern to be obtained from the same area of the sample.
Other surface investigative techniques involve Scanning Tunnelling Microscopy
(STM)35 often used alongside Auger analysis of the electrons emitted, thus
providing comprehensive surface analysis. Recently, the powerful technique of
Penning Ionisation Electron Spectroscopy (PIES)36 has been applied to this field.
This is a very surface sensitive technique in which the electrons do not penetrate
103
deep into the film.
Having obtained relevant film dimensions and an idea of the film morphology then
it is clearly useful to study the film orientation. One of the most commonly used
techniques is that of linear dichroism (LD).37 This makes use of an anisotropic
absorption of light by the film. If the electric vector of the polarised light is
altered, then the absorption line will vary by an amount dependent on that electric
vector and the transition dipole moment of the particular molecule. Such readings,
if varied, can ultimately give rise to the various orientations of the molecule. This
is clearly useful for highly polar molecules where the transition dipole moment is
particularly well defined. Where the above dipole moment is not particularly
defined, then the infra-red absorption bands are more commonly used as these are
generally more distinct and easily categorised - though they are not as susceptible to
orientational changes. Other useful techniques for studying molecular orientation
are resonance Raman spectroscopy38 and electron spin resonance spectroscopy39 -
though the latter of these techniques requires specific film requirements which must
be met. Electron spin resonance is dependent upon the presence, to some degree,
of free radical species and as such has been used in the study of LB film of TCNQ,
and in particular conducting TCNQ/cationic films. As was discussed earlier, these
systems are radical anion/cation systems.
3.6 Applications of LB Films
The proposed applications of LB films are many and varied and cover the range
from electronics to biology. Various applications shall be discussed, though it must •
be stressed that there is a large amount of fundamental research needed before any
real use will be found. For example, although there are many LB films that can be
produced with a good deal of stability, to be able to do this reproducibly and to
predict the stability of unknown materials, is still proving elusive. It is also not
104
certain that the technology exists to be able to fabricate the amounts of LB films
needed to justify a commercial application, even if one could be proposed with
confidence. However, although there is a long way to go, the number of proposed
applications is numerous, and the field is expanding at such a rate as to give some
cause for optimism.
To show both the fundamental and cross-disciplinary nature of LB film research,
various biological film applications can be illustrated. LB multi-layers bear a
striking resemblance to biological membranes. Such membranes essentially consist
of fatty molecules - in fact phospholipids - together with a certain amount of
carbohydrate and protein. Because such lipids are amphipathic then they tend to
spontaneously orientate as an organised assembly. The structure of such an
assembly is directly comparable to an LB film and Yoshikawa et al used such a film
as a model system to study biological membranes.40 Essentially, the system
investigated attempted to "harvest" light energy which was then, by a process of
many electron transfers, used to produce oxygen. The whole process was carried
out at the molecular level. The light harvesting, energy transfer and subsequent
reactions, were all successfully simulated artificially by using monomolecular layer
assemblies. The assemblies consisted of donor and acceptor molecules interspersed
with a harvester and sensitiser. Overlaps in the emission spectra of each enabled
energy transfer via electron transfer to take place.
It has already been mentioned earlier that the adducts studied in this work may find
uses in non-linear optical applications. The LB film technique has prompted great
interest in this area because of its ability to control (to a certain degree) the
molecular orientation and structure of the films. The key to many applications in
this area is the non-centrosymmetric nature of the system chosen for study. Very
few single crystals satisfy this requirement and so Z or X-type LB films which are
of a non-centrosymmetric nature have generated much interest. However, X or Z-
105
type films are generally found to be of poorer quality than Y-type structures for
some reason. However, second harmonic generation has been reported in
monomolecular layers42 though the non-linear coefficient was not particularly large.
Such observations have also been seen in non-centrosymmetric multilayers.43 There
have been reports of large second order coefficients using Y-type deposition. In
one case the non-centrosymmetric structure results from a unique herringbone
arrangement of the chromophores.44 Other examples have relied on the principle of
depositing alternate layers of different materials45 - in this case the dipole moments
of adjacent molecules would not cancel and the non-linear optical properties would
be maintained.
Much work in the past has been directed towards the potential use of LB films as
very thin resists.46 As integrated circuit technology moved more and more to
smaller and smaller circuit elements - as faster speeds and larger memories are
sought - so the need for much better resolution is required. Thus, photolithographic
techniques are being replaced by X-rays, electron and ion-beams. Extremely thin
resists are one way in which the problem of scattering of substrate material around
the point of impact when ion-beams are used can be reduced. The use of such thin
resists require defect levels of negligible proportions which the LB technique has
not yet shown itself to be reliably capable of.
LB films are generally excellent insulators, eg CO -tricosenoic acid. They have
therefore been the subject of much research towards their potential use in field
effect semiconductor devices47 and have been successfully used for active insulators
on compound semiconductors.48 However, because very many of the LB films
studied have very low temperature tolerances, many manufacturers have been
reluctant to use them. While there do exist films with very high temperature
tolerances - for example aromatic or polymeric based materials - these are often
films of much poorer quality, and so once again there appears to be the unfortunate
106
"trade-off' between film stability and quality.
However, there are grounds for optimism in the future of LB film research. Whilst
admittedly there is no tangible evidence of electronic application after substantial
funding and effort during the past few years, the unique ability to exercise a degree
of control over the film development using this technique is encouraging.
Certainly, the scope for producing suitable organic amphiphiles is endless and so it
can only be a matter of time before the molecule is tound that can provide the easily
transferred, defect-free, stable film sought.
Papers presented at the bi-annual conferences on Langmuir-Blodgett films have to
date been published as special volumes of Thin Solid Films. The proceedings of the
last conference in France49 brings the reader up-to-date with the current state of the
art.
107
3.7 REFERENCES
1 G G Roberts, Adv Phys, 3 4 , 475, (1985)
2 I R Peterson, J Phys D Appl Phys, 23, 379, (1990)
3 D Tafor, J Colloid Interface Sci, 75, 240, (1980)
4 T Terada, R Yamamoto and T Watanabe, Sci Pap Inst Phys Chem Res Jpn, 23,173, (1934)
5 B Franklin, Phil Trans R Soc, 64, 445, (1774)
6 Lord Rayleigh, Proc R Soc, 47, 364, (1890)
7 A Pockels, Nature, 43, 437, (1891)
8 I Langmuir, J Am Chem Soc, 39, 1848, (1917)
9 K B Blodgett, J Am Chem Soc, 57, 1007, (1935)
10 H Kuhn, D Mobius and H Bucher, Physical Methods in Chemistry, I , 577,(1972)
11 A Barraud, C Rosilio and A Ruaudel-Teixier, J Colloid Interface Sci, 62, 509, (1977)
12 B Tieke, G Wegner, D Naegele, and H Ringsdorf, Angew Chem, Int Ed Engl, 15 , 764, (1976)
13 M F Daniels, O C Lettington, and S M Small, Thin Solid Films, 99, 61, (1983)
14 T J Lewis, D M Thylor, J P Llewellyn, S Salvagno and C J M Stirling, Thin Solid Films, 133, 243, (1985)
15 G L Gaines, Insoluble Monolayers at Liquid-Gas Interfaces, Interscience, New York, (1966)
16 R H Tredgold, Thin Solid Films, 152, 223, (1987)
17 A Barraud, P Lesieur, A Ruaudel-Teixier and M Vandevyver, Thin Solid Films, 133, 125, (1985)
18 M Fujiki and H Thbei, Synthetic Metals, 18, 815, (1987)
19 S Baker, M G Petty, G G Roberts and M V Twigg, Thin Solid Films, 29, 53, (1982)
20 R H Tredgold, S D Evans, P Hodge, R Jones, M G Stocks and M C J Young, Brit Polym J, 19, 397, (1987)
21 P S Vincett, W A Barlow, F T Boyle, J A Finney and G G Roberts, Thin Solid Films, 60, 265, (1979)
108
22 A Barraud, M Florsheimer, H Mohwald, J Richard, A Ruaudel-Teixier and M Vandevyver, J Colloid Interface Sci, 121. 491, (1988)
23 O Albrecht, Thin Solid Films, 178. 563, (1989)
24 T Nakamura, M Matsumoto, F Tahei, M Tanaka, T Sehiguchi, E Manda and Y Kawabata, Chem Lett, 709, (1986)
25 K Fukuda, H Nakahara and T Uato, J Colloid Interface Sci, 54, 430, (1976)
26 M Francon, Rev Opt 2 (1952)
27 JO Birzer and H J Schulzer, J Colloid Polym-Sci, 264, 642, (1986)
28 B Rothenhansler, C Duschl and W Knoll, Thin Solid Films, 159. 323, (1988)
29 M Pomerantz and A Segmuller, Thin Solid Films, 68, 33, (1980)
30 I R Peterson, G J Russell and G G Roberts, Thin Solid Films, 109. 34, (1983)
31 I R Peterson, G J Russell, D B Neal, M C Petty, G G Roberts and T Ginnais, Philos Mag B ,M , 71,(1986)
32 V Vogel and C Woll, J Chem Phys, M, 5208, (1986)
33 J B Lando and J E Hansen, Thin Solid Films, 180. 141, (1989)
34 R A Hann, S K Gupta, J R Fryer and B L Eyres, Thin Solid Films, 134. 35, (1985)
35 J H Coombes, J B Pethica, M E Welland, Thin Solid Films, 159. 293, (1985)
36 Y Harada, Surf Sci, 158. 2455, (1985); H Ozaki, Y Harada, K Nishiyama and M Fujihara, J Am Chem Soc, 109. 950, (1987)
37 K Ogawa, H Yonehara and E Maekawa, Thin Solid Film, 210. 535, (1992)
38 M Vandevyver, A Ruaudel-Teixier, L Brehamet and A Barraud, Thin Solid Films, 99. 41,(1983)
39 M Vandevyver, A Barraud, A Ruaudel-Teixier, P Maillard and C Gianotti, J ColloidInterface Sci, 85, 571, (1982)
40 K Yoshikawa, M Mahino, S Nakata and T Ishii, Thin Solid Films, 180. 117, (1989)
41 M Fujihira, M Sakomura, T Kamei, Thin Solid Films, 180. 43, (1989)
42 I R Girling, N A Cade, P V Kolinsky and C M Montgomery, Electronics Letts, 21, 169,(1985)
43 I R Girling, N A Cade, P V Kolinsky, J D Earls, G H Cross and I R Peterson,Thin Solid Films, 132. 101 (1985)
109
44 G Decher, B Tieke, C Bossard and P Gunther, J Chem Soc Chem Commun,933, (1988)
45 I Ledoux, D Joise, P Frenaux, J P Piel, G Post, J Zyss
T McClean, R A Hann, P F Gordon and S Allen, Thin Solid Films, 160, 217,(1988)
46 M Vekita, H Awaji, M Murata and S Mizuuma, Thin Solid Films, 180, 271,(1989); H Kato, M Thwata, S Morita and S Hattori, Thin Solid Films, 180, 299,(1989)
47 G L Larkins Jr, C D Fung and S E Richert, Thin Solid Films, 180, 217, (1989)
48 J P Lloyd, M C Petty, G G Roberts, P G Lecomber and W E Spear, Thin Solid Films, 99, 297, (1982)
49 Thin Solid Films, Vols 210 - 211, (1992)
110
CHAPTER 4: LANGMUIR - BLODGETT FILMS - EXPERIMENTAL
This chapter discusses the design and operation of the Langmuir trough used in this work. The
more esoteric experimental considerations regarding isotherm recording, film stability, etc are
discussed in Chapter 5.
4.1 The Joyce - Loebl Monolayer Coating Unit
This commercially available trough is of proven design and was originally
developed by scientists at ICI and Durham University based on the idea of Blight et
al1. It is used successfully by many academic and industrial research groups
throughout the world.
The two essential features necessary in the operation of a Langmuir Trough -
isotherm recording and monolayer removal - are fully automated. All mechanical
units are made from corrosion resistant material. The main features of this, and
indeed any other Langmuir Trough, are shown schematically in fig 4.1.
4.2 The Compression System
Traditionally, the container holding the liquid subphase forms an integral part of the
compression system. In such a system, it acts as part of the boundary between the
floating monolayer on the subphase and the clean water (monolayer free) beyond.
Consequently, there is the need for incorporation of a suitable seal to limit any
potential film leakage.
Ill
ELECTROBALANCE C OM PARATOR
BARRIER>v MONOLAYER^•liiiiiiiniuiiiuinii✓✓ Q
SUBPHASE Itg , purified vafer)
-7-- ////
B A R R IE RH O T O R
-TROUGH
Figure 4.1 Schematic diagram of the essential features of a Langmuir Trough
112
An important feature of the Joyce - Loebl trough is the incorporation of a constant
perimeter PTFE coated glass fibre barrier to define the working area. This system
keeps the problems of film leakage (mentioned above) and contamination to a
minimum. The design of the barrier is outlined in fig 4.2.
The compression barrier is a belt of width 2cm, and is held in place by a system of
six PTFE rollers which are in turn secured by two mobile overarms. By means of a
highly geared motor the mobile overarms are moved symmetrically inwards or
outwards. At all times the barrier is kept taut and thus the size of the film area can
be carefully controlled. The study of the film in both compression and expansion
mode is possible at a range of speeds.
4.3 Surface Pressure Measurement
The Wilhelmy Plate technique used to monitor the surface pressure is very reliable
and can be used accurately for a range of surface pressures. A sensitive
microbalance attached to a sensor (typically a piece of Whatman Grade 1
chromatography paper) in the subphase measures the differential pressure. This is
shown schematically in fig 4.3.
As is shown, the plate is semi-immersed in the subphase and is attached to a
microbalance directly above it via a thread. Hence, counterbalancing forces on the
microbalance can be directly linked to the surface pressure provided the
microbalance is sensitive enough to cope with the small forces involved.
113
Figure 4.2 Molecules deposited within an area defined by PTFE barrier as shown
a) Maximum Area
b) Minimum Area
CD
115
Figu
re
4.3
Rep
rese
ntat
ion
of Su
rfac
e Pr
essu
re
Mea
sure
men
t
The forces acting on the plate are gravity and surface tension downwards, and
buoyancy due to displaced water upwards. Thus, for a plot of the dimensions
shown - length '1', width 'w ' and thickness ’t*, immersed to a depth 'h' with a
contact angle 0 (fig 4.3), the net downward force is given by
where & is the surface tension; 'g ' the gravitational constant and ' p* is the density
(j£ = paper density; = subphase density). The surface pressure 'x ' is regarded
as being the reduction in surface tension ocurring when the subphase is covered by
a monolayer - ie it is the expanding pressure of the monolayer opposing the natural
contracting tension of the subphase surface.
Eq (7)
Thus,
AY
where Y 0 = surface tension of the clean liquid, Y* = surface tension of the
subphase covered with a monolayer.
Hence
Eq(8)
assuming that the contact angle, 0, is zero (ie cos0 = 1).
116
Thus, if the plate is of negligible thickness and width 1cm and ignoring the upward
force due to the displaced monolayer, then
F = 2 &Y Eq (9)
ie, the weight measured in mg is equal to twice the surface pressure measured
in m N-m '1.
4.4 Trough Operation
The detail given below is quite general, and is not specific to any one material or
indeed any subphase.
4.4.1 Cleanliness
The need for scrupulous cleanliness has already been shown and its importance
cannot be stressed too highly. Many of the problems encountered in monolayer
work are due to the presence of impurities and other "surface active" contaminents.
Rigorous cleaning procedures and the use of highest purity solvents available are
essential, as well as carrying out experiments in the cleanest environment possible.
The trough was cleaned at least three times weekly using alternate washings of
dichloromethane or chloroform, iso-propyl alcohol and high purity water. The
barrier and rollers were also cleaned at the same time, using the same procedure.
Although the system incorporated a fan to presumably keep the air around the water
subphase circulating, it was found that this introduced severe turbulance on the
water subphase and consequently was not used. Indeed, the whole working system
was mounted on an anti-vibration table.
117
4.4.2 Trough Calibration
Before use, all separate instruments - pH meter, microbalance and chart recorder
were calibrated in accordance with their respective manufacturers instructions, and
recalibrated at regular intervals.
4.4.2.1 Calibration of Surface Area
For reproducible, accurate isotherms to be obtained then the abscissa of the X-Y
chart recorder must be calibrated. To do this, knowledge of the maximum and
minimum areas within the constant perimeter barrier is required. The approximate
values of these are 1000cm2 and 100cm2, though all researchers should determine
the exact value of these for themselves. Because the speed at which the barrier is
moving prior to attaining the maximum and minimum areas affects exactly when the
microswitches trip, the speed used for area calibrations should be that at which the
isotherm measurement takes place.
4.4.2.2 Calibration of Surface Pressure
When the microbalance has been correctly calibrated by the use of standard
weights, then the surface pressure is directly related to the dimensions of the
Wilhelmy plate as discussed previously at 4.3. Thus, the ordinate of the X-Y
recorder can be most readily calibrated by the addition of known weights to the
microbalance.
118
4.5 The Subphase
Most of the work carried out on Langmuir Troughs requires an aqueous subphase.
The purest quality water available is used. In this work a Millipore "Milli-Q" water
purification unit was employed feeding water of 18mQ resistance directly into the
trough as and when required. This system makes use of an initial reverse osmosis
stage followed by a system of "polishing" filters which ultimately remove
contaminants at the biological level. When adding divalent metal ions to the
subphase (to aid certain film stability, and discussed earlier), then these must also
be of the purest grade available. Similarly with the addition of acid or alkali to
raise or lower the subphase pH. It is advisable to change the water - cleaning the
trough at the same time - at least every other day, and certainly after the weekend,
or whenever the system may not have been used for some time.
4.6 Subphase Surface Cleaning
Obviously, for reproducible results to be obtained, then the level of surface
contamination should be kept to an absolute minimum. It is therefore necessary to
clean the subphase surface prior to the spreading of the monolayer material. Such
cleaning is routinely done by attaching a clean glass pasteur pipette (preferably new)
to a suitable suction pump and then sucking the surface clean. Depending on the
mains pressure then a simple water pump will suffice, or if this is not available then
a pump capable of taking a mixture of both air and water is necessary. Once the
surface has been cleaned then its quality must be checked. Checking for the degree
of contamination present is easier if the surface is cleaned at minimum area (ie fully
closed), then opening the barrier and closing slowly. Any appreciable increase in
the surface pressure is indicative of surface contamination and the process should be
repeated. Ideally, if the subphase is totally clean then there will be a zero increase
in the surface pressure.
119
This technique has recently been criticised by Albrecht2 who pointed out that any
system where the barrier cannot close to give an absolute zero, will still have an
appreciable area left which, on compression, could show surface contamination.
The way to overcome this may be to spread and compress a well characterised
amphiphile (for example, stearic acid) and observe any deviations from the standard
isotherm.
4.7 Monolayer Material Preparation and Spreading
A standard solution (known concentration) of the desired material must first be
produced in a suitable solvent. Solvents commonly used in Langmuir-Blodgett
work are shown in Table 4.1
SOLVENT Mpt (°C) Bpt (°C) SOLUBILITY g/1000g H20
n-hexane - 94 69 0.01cyclohexane 6.5 81 0. 07benzene 5.5 80 1.8chloroform - 64 61 8diethyl ether - 116 35 75
Table 4.1 Common solvents, with relevant data, for Langmuir-Blodgett film preparation
The minute volumes of material in solution needed, can be deposited onto the
subphase by means of a micrometer syringe - of which several commercial varieties
are available. On occasion mixed solvent systems are used, particularly where one
solvent "spreads" differently from the other.
120
When producing such solutions, extreme accuracy is required in order to obtain
reproducible results - the mass of adduct dissolved in the solvent was in this work
measured to one hundredth of a milligram. The amount of solvent used was
typically one millilitre - delivered by a Gilson pipette. The amount of solution then
deposited onto the subphase was known to the nearest ten microlitres. The amount
of material deposited on the subphase should be calculated to give a compressed
monoiayer w'hen the barrier is half-closed between the minimum and maximum
areas. The material is added dropwise with the syringe tip head just above the
subphase. Once deposited, then time is allowed for any excess solvent to evaporate
- approximately 5 minutes is generally enough.
4.8 Temperature of the Subphase
Changes in the properties of a monolayer forming material can be due to-increasing
or decreasing the temperature of the subphase. The temperature can be altered by
incorporating a heat exchanger into the trough, and measuring the temperature with
a suitable probe. It must again be stressed that the heat exchanger and temperature
probe should be thoroughly checked for cleanliness if they are to come into contact\
with the subphase.
4.9 Compression of the Monolayer
Before the compression of the monolayer can begin, the microbalance must be
zeroed. It is routinely straight forward to do this by adding or removing weights
from the microbalance pan to give a coarse zero and then finalising with the zero
control key. Cleaning the subphase will change any zero reading obtained as it
removes water from the trough, so all cleaning should be carried out before the
system is zeroed. It is more usual for the compression to be carried out slowly - the
121
typical time scale for an isotherm measurement being about 1 5 -2 0 minutes.
Once the essential features of the isotherm have been recorded, the barrier is
returned to its original open position. This avoids the collapse of the monolayer if
it is compressed too far. However, when studying a novel material for the first
time, the collapse pressure is usually recorded.
4.10 Control of Surface Pressure
To facilitate monolayer removal to the substrate, it is useful to be able'to control the
surface pressure of the monolayer. A negative feedback loop provides the facility
to maintain the system at a constant pressure - compensating for removal of material
by closing the barrier by an appropriate amount. A feedback time constant puts
additional damping into the feedback loop and can thus change the response time of
the system.
4.11 Transfer of Monolayer
Much of the details of the monolayer removal - in particular the various ways in
which a materials may be deposited on a monolayer - have been mentioned in the
previous chapter. The number of layers that may be required to be deposited, the
drying time between successive layers, the speed of the deposition and the depth can
all be selected and controlled using this system. During the deposition process, all
vibrations should be kept to an absolute minimum and the system is often mounted
on an anti-vibration chamber, and preferably in a basement room.
The deposition process is most conveniently monitored using a two-channel
recorder to measure both the surface pressure and surface area against time. If the
system is behaving correctly, then a decrease in area should be observed as material
122
is being removed from the subphase, due to the negative feedback loop maintaining
the surface pressure at the predetermined value. If the area remains steady then this
would indicate that there is no material being transferred.
Photographs 1 and 2 show various features of the Langmuir Trough used in this
work, and the layout of the Langmuir-Blodgett laboratory at the Health and Safety
Executive.
123
Photograph 1 : The Langmuir-Blodgett Laboratory
124
Photograph 2 : Details of the Compression System and
Dipping Head.
\ XS
4.12 REFERENCES
1 L Blight, C W N Campbell and V J Kyte; J Colloid Sci, 20, 393, (1965)
2 O Albrecht, Thin Solid Films, 178. 563, (1989)
126
CHAPTER 5: LANGMUIR - BLODGETT FILMS - RESULTS AND DISCUSSION
A systematic study of the film forming potential of the donor-x-acceptor zwitterionic adducts
was carried out in collaboration with scientists at the Health and Safety Executive Research and
Laboratory Services Division in Sheffield. Very briefly, the study concentrated on three main
areas:
a) The behaviour of the materials on the water subphase
b) The deposition of the materials as LB films
c) The characterisation of the films by ultraviolet/visible spectroscopy
The results and discussion of each of these areas are discussed in turn below.
5.1 Langmuir Film Studies on R(4)Q3CNQ
The molecular structure of such donor-x-acceptor adducts leads one to suspect that
they may form an insoluble, monomolecular film at the air water interface. The
fact that these adducts are sparingly soluble in many organic solvents, and that
many such solvents are not suitable for spreading on water, was not a serious
problem as the materials are soluble in dichloromethane. Thus, solutions of the
materials could be prepared by weighing out the appropriate adduct and then
dissolving in an appropriate amount of solvent delivered by a Gilsen pipette. The
concentration of the resultant solution was typically lmg ml"1. All glassware and
the spatula were thoroughly cleaned beforehand and a new pipette was used each
time in order to keep any external contamination to a minimum.
An appropriate volume of the solution was then delivered to the water subphase,
dropwise, by means of a micrometer syringe. Between each drop time was allowed
for the solvent to evaporate and for the materials to achieve free association with the
127
subphase. Once all the solvent had been added and allowed to evaporate - typically
after a period of ten minutes - the monolayer was compressed and the surface
pressure vs area per molecule characteristics were studied.
The Joyce-Loebl Langmuir Trough 4 System will calculate the area per molecule if
certain parameters are known. These are the maximum and minimum areas
enclosed by the barrier, which are set internally during the calibration routines.
The concentration of the solution added to the subphase is also calculated, knowing
the volume of solution added to the subphase. Thus the area occupied per molecule
can be calculated.
The shape and form of the isotherms for the range of materials studied are shown in
figures 5.1 to 5.8. The scale of the area per molecule axis is chosen by the
instrument software, explaining certain of the anomalies in scaling along the area
axis. Table 5.1 shows the values of the area molecule for each adduct at
25 mNm'1. It was noted that there was a definite change in molecular arrangement
occurring at the alkyl chain lengths greater than 15 carbon atoms as is clear from
the sudden increase in the area per molecule. In view of the molecular geometry of
the adduct, the area values are quite small and a more detailed discussion of the
structure of the Langmuir films is offered later in the chapter.
Figures 5.9 and 5.10 show successive compressions and expansions of the
Langmuir film and it can be seen that the degree of hysteresis on successive
compression and expansion is minimal. The problem of film hysteresis has been
acknowledged for some time.1 Very simply, if the molecule is compressed and then
expanded, then in a fully dynamic system, it would be expected that there would be
no change in the pressure/area profile.
128
E 8cr..cU ; >
4 9 1 \i /Surf sc? f r e - u r s IitH/rt 38
10-j
V ---------------- :---------------- :---------------- :--------------------------------- :0 2! 42 84 35 !98
Arsa/Ho I ecu is (ftngsbroftA2)
Figure 5.1 Isotherm showing change in surface pressure (mNm*1) against Area per molecule (A*) for C8(4)Q3CNQ
129
Surface IPpp<;c.npp jttN/rt “ ‘38 \
2lJ -
Jrj 1ii • *r .
! *\f t • vU ------------------- :------------------- :-------------------:---- :-------------:-------------------,
9 299 400 899 899 1990 E 9Fiift area (caa2)
Figure 5.2 Isotherm showing change in surface pressure (niNm-1) against Area per molecule (A1 ) for C9(4)Q3CNQ
E 0 50-
40- Surface Pressure tiN/n 30 -
20J
10
0 -
0 25
' \\• \
51, 28Ar̂ s/Molec-uIs
102 strottA2)
127 E 0
Figure 5.3 Isotherm showing change in surface pressure (mNm'1) against Area per molecule (Aa ) for C11(4)Q3CNQ
131
Z 8c-8 •>
40 iSurface j Pressure ! riH/n 38:
42 t'4 f'S0 J
Hraa/sic-1 ecu I s (AnsstrG:H.A2)
Figure 5.4 Isotherm showing change in surface pressure (mNm*1) against Area per molecule (a ) for C13(4)Q3CNQ
132
i.i j
c nL . U
c0 -vfb '
43P itr ts c ?i l i i T-2!
Pressurs ril/n 33
23
i n ;
2 ! 42 S4 85 !Q6 E 3Arss/Ho I sen is (r.ngstro!'.A2)
Figure 5.5 Isotherm showing change in surface pressure (mNm"1) against A rea molecule (A ) for C14(4)Q3CNQ s Area per
133
Figure 5.6
9 38 75 113 <50 \m i 6Ares/Ho 1 ecu I e (Anss trortA2)
Isotherm showing change in surface pressure (mNm ) against Area per molecule (A*) for C15(4)Q3CNQ
Figure 5.7 Isotherm showing change in surface pressure (mNm'1) against Area per molecule (A* ) for C16(4)Q3CNQ
30 SO 90 120 130 E 0Ares/HoIecu Ie (nngsirGn*2)
Figure 5.8 Isotherm showing change in surface pressure (mNm*1) against Area per molecule (A*1) for C20(4)Q3CNQ
E 8
58 :
4 9 JSurface j pressure I «H/« 38 -I
i;t
70 - ’K“w i
• n•0- V-• 'J
0 ------------------------------
\9 24 47 7! ?4 US Z 0
fra a/Ho i ecu I e if n=a iron A2)
Figure 5.9 Four successive compression/expansion cycles showing negligible film hysteresis - C11(4)Q3CNQ
137
Figure 5.10 Further compression/expansion cycles showing negligible film hysteresis C16(4)Q3CNQ
R ( N o . o f c a r b o n a t o m s ) A R E A P E R M O L E C U L E ( A 2 )
C 8 3 2
C 9 30
C I O 3 4
C l l 3 0
C 1 2 3 0
C 1 3 3 1
C 1 4 3 1
C 1 5 4 0
C 1 6 4 4
C 1 8 4 2
C 2 0 4 8
Table 5.1 Area per molecule data for R(4)Q3CNQ adducts at 25 mNm'1
139
This is because, on compression, each molecule would be expected to organise into its
preferred orientation and on expansion to return to a state equivalent to its original integrity.
While molecules can obtain a preferred orientation on compression, they can be preferentially
influenced by neighbouring molecules on successive expansions and compressions and thus
retain an aggregated state. A routine compression/expansion experiment of this type was
therefore carried out on each zwitterion to see whether such aggregation is present.
If the materials are to be successfully transferred to a solid substrate and fabricated as LB films
then the compressed monolayer should be stable at the chosen deposition pressure. Thus, the
barrier can be held at a particular surface pressure and the film area monitored with time. If
the material is undergoing structural reorientation or if the film is collapsing and dissolving
into the subphase then a decrease in the film area will be observed as a consequence of the
feedback system maintaining the surface.pressure at the predetermined value. Figures 5.11 to
5.13 show that the films are stable, with a 0.1% decrease in film area occurring over a period
of one hour.
This work has shown that these materials form monomolecular layers at the
air-water interface. The isotherms obtained are not of the classic three phase shape shown
in fig 4.3 which is normal for long chain stearic and other fatty acids, though this is not
really surprising. The molecular features - long alkyl chain and small, typically carboxyl, head
groups - which often lead to defined phase changes in more condensed films, are almost all
missing in long chain adducts of R(4)Q3CNQ. For example, such a bulky head group does
not tend to favour efficient packing and maximum cohesion between hydrocarbon tails,
nor is the presence of two polar moieties in the chromophore favourable. Also, the attraction
of one polar group for the aqueous subphase would have to be overcome before the most
stable state is reached. This would be a slow process and distinct gaseous-liquid-solid phase
140
l Iapsed tine (secs5
Figure 5.11 Stability of C8(4)Q3CNQ film at 25 mNm'
£ 91969 -
li
300i 1 ift
<2 *2)5301
avu-l
208 -!
0 :6
EUpScd tine (ice-)
Figure 5.12 Stability of C16(4)Q3CNQ film at 25 m N m '1
142
Figure 5.13 Stability of C11(4)Q3CNQ film at 25 mNm'1
changes would not be expected to be observed.
The effect of pH and added salt (say Cd2+) was also found to be minimal. The
change in film characteristics on changing the pH is again more prevalent in more
"conventional" films as is the effect of adding a divalent metal ion to aid the film
cohesion.
5.2 Langmuir-BIodgett Films of R(4)Q3CNQ adducts
The transference of material from the condensed monolayer to a suitable substrate
can be attempted once a knowledge of the materials' behaviour outlined in 5.1 has
been fully investigated. The mechanics of the LB technique have been fully
described previously.
The adducts were deposited onto glass (silica) substrates which had been treated to
render the surface hydrophilic. This was done by the following technique involving
several successive washes in water (18 MQ milli-Q), iso-propylalcohol and
dichloromethane. The slides were then left to soak overnight in a concentrated
solution of sodium hydroxide. Before use, each freshly prepared slide was (always
handling with tweezers) washed in copious quantities of ultra pure water and blown
dry with filtered nitrogen ("white spot").
The condensed film was held at 25 mNm'1 and the slide, held in the dipping head
(shown in photograph 2, Chapter 4) was passed vertically through the subphase
from air to liquid, at a rate of 5 mm/min*1. After travelling a predetermined
distance into the subphase, the slide was then removed at the same rate. This
procedure was then repeated several times, thus forming a stable film on the slide..
As described in section 4.4, the transfer ratio 'y can be determined to establish
whether the resultant film is X,Y or Z-type. Each stroke down or up constitutes
one possible layer transferred. Selected transfer ratios and deposition profiles are
144
shown in table 5.2 and 5.3 and these results are shown in bar graphs 5.14 and 5.15.
It is clear that neither pure X,Y or Z-type deposition occurs, rather that the films
are predominantly Y-type with material being transferred on both the up and down
stroke. As a consequence of the hydrophilically treated slide, no material is
transferred on the first downstroke. This is because the surface of the substrate,
being hydrophilic, attracts water, and the first immersion of the slide into the
subphase results only in the wetting of the slide (fig 5 .16(a)). On removal, the
hydrophilic head is deposited on the slide as a consequence of the water substrate
attraction shown (fig 5.16(b)). The substrate surface is then rendered hydrophobic
and thus on the next immersion the hydrophobic tails will be attracted to the
substrate. This leads, after successive passages through the subphase, to a head-to-
head, tail-to-tail structure. Thus, films can be produced with a certain degree of
order.
The next portion of the work involved the further characterisation and study of the
films by a variety of techniques with a view to establishing their structure more
fully.
5.3 Characterisation of LB films of R(4)Q3CNQ adducts
It has already been shown that the transfer ratio can give some indication as to
whether or not one monolayer at a time is being transferred. The integrity of the
films can be assessed in other ways as mentioned in chapter 3 and, to illustrate one
method, an absorbance vs number of layers plot for C15(4)Q3CNQ is shown in
fig 5.17. This essentially straight line plot is indicative of good, reproducible
monolayer transference.
145
L a y e r N u m b e r T r a n s f e r R a t i o ( y j )
1 0
2 1 . 0 9
3 0 . 7 6
4 1 . 0 9
5 0 . 7 6
6 1 . 0 6
7 0 . 7 6
8 1 . 0 4
9 0 . 8 2
1 0 0 . 9 9
1 1 0 . 7 6
1 2 0 . 9 9
1 3 0 . 9 9
1 4 1 . 0 0
1 5 0 . 9 9
Table 5.2 Selected transfer ratios for C16(4)Q3CNQ
146
L a y e r N u m b e r T r a n s f e r R a t i o ( ~|fj )
1 0
2 0 . 6
3 1 . 1
4 0 . 6
5 0 . 6
6 1 . 1
7 C . 6
8 0 . 6
9 1 . 1
1 0 0 . 6
1 1 1 . 1
1 2 0 . 6
Table 5.3 Selected transfer ratios for C14(4)Q3CNQ
147
Figure 5.15
Com
puted transfer
ratios for
C16(4)Q
3CN
Q
T r a n s f e r R a t i o
ro r
Qj r
CJl
CJ)
O CD
CO
ro
03 •
cn
148
Figure 5.14
Com
puted transfer
ratios for
C14(4)Q
3CN
Q
T r a n s f e r R a t i o
oo ro cu
■f— TCJl 01“i — r
CD ID OT
ro"1
ro
CO
r-0)*<CD1
O
CJ1
CJ)
CD
149
' \ V/ \ x 5uSPHfls£
- J
F ig u r e 5 .1 6 D e p o s i t i o n p ro c e s s s h o w in g :
a) 1st downstroke (no transfer)b) 1st upstroke (transfer)
The resultant substrate surface is hydrophilic
Figure 5.17
Absorbance
vs No
of layers
for C
15(4)Q3C
NQ
A bsorbance
zo
r~O)*<ro~3LO
o
o
151
Figure 5.17
Continued... A
bsorbance vs
No of
layers for
C14(4)Q
3CN
Q
Absorbanceo
C J
o
o
ro
152
The films were also studied using uv/vis spectroscopy. Selected spectra are shown
in figures 5.18 to 5.26. It is apparent that the band is considerably sharpened as
compared to the solution spectra and shifted to shorter wavelength, with a further
shift to shorter wavelengths as the chain length increases. The measured A max
(nm) values are shown in table 5.4. A fuller discussion as to why this should be
observed is given in the later part of this chapter.
A study of the film morphology and structure was made using Transmission
Electron Microscopy (TEM). The procedure used to study the films was that first
described by Zingsheim2, and which has been used by many researchers since
then3. This particular microscopic technique is used because TEM requires the use
of very thin substrates. The particular material is deposited as an LB film onto a
glass slide coated with A120 3 (ie aluminium which has been allowed to slowly
oxidise in air). The film is then "flashed” with carbon, thus enabling the area
covered by the LB film to be visualised. Once this has been done, the slide is then
lowered into a solution of mercury (II) chloride by some suitable instrument. It is
critically important when carrying out this stage that the slide is lowered into the
mercury solution at the correct angle to ensure the solution attacks the aluminium
and does not attack the film through the carbon. The LB film can then be
transferred to microscope grids and is then ready for examination. The advantage
of studying these films by TEM is that the transmission electron microscope allows
electron diffraction patterns to be obtained from the same area that is under
observation. Thus, the film texture and the underlying crystallographic pattern can
be obtained from the same part of the film.
153
9ee .— s.ie i
-e .0210 . 021
Figure 5.18 UV/VIS spectra for a 6-layer LB film of C8(4)Q3CNQ
154
7*0*
0 0-
0 007
4
Figure 5.19 UV/VTS spectra for a 3-layer LB film of C9(4)Q3CNQ
.055
0 - ?
02?
0 0>:
0 . 0 0 3
0 010
155
see
Q 112
e 000 ■0 . O0v
Figure 5.20 UV/VIS spectra for an 8-layer LB film of C11(4)Q3CNQ
T O O
Figure 5.21 UV/VIS spectra for an 8-layer LB film of C12(4)Q3CNQ
157
see
Figure 5.22 UV/VIS spectra for a 5-layer LB film of C13(4)Q3CNQ
158
see
e .si*£
Figure 5.23 UV/VIS spectra for a 4-layer LB film of C14(4)Q3CNQ
159
5 * 0 see 700 9ee
- 9.0316
0 01S8
Figure 5.24 UV/VIS spectra for a 5-layer LB film of C15(4)Q3CNQ
160
Figure 5.25 UV/VIS spectra for an 8-Iayer LB film of C16(4)Q3CNQ
<5 04“- —
1
0 0340 “
— S«
L
- 0 0“-l<
8 . 805-
Figure 5.26 UV/VIS spectra for a 4-Iayer LB film of C20(4)Q3CNQ
162
R ( N o . o f c a r b o n a t o m 9 ^ m a x ( n m )
C 8 6 1 6
C 9 6 1 7
C I O 6 1 6
C l l 6 1 6
C 1 2 6 1 3
C 1 3 6 1 5
C 1 4 6 1 5
C 1 5 5 6 1
C 1 6 5 6 5
C 1 8 5 6 5
C 2 0 5 6 8
Table 5.4 Variation in ^ max for LB films of R(4)Q3CNQ adducts
163
Using a 100 KeV microscope the film morphology and corresponding diffraction
patterns were observed. These are shown in figures 5.27 to 5.29. This data is for
an 8-layer film of C16(4)Q3CNQ. The diffraction patterns are very encouraging,
showing, as they do, excellent structural order. Such hexagonal patterns are
characteristic of the head-to-head, tail-to-tail, etc structure proposed on the basis of
the observed predominantly Y-type deposition. There is also a certain degree of
"twinning" of the diffraction pattern. This would appear to suggest that there is a
separate orientational regime for the "head" as opposed to the "tail" - this is
expanded in the next section in which a more detailed description of the film
structure is proposed and described. Examination of the X-ray spectrum of each
area studied (Figs 5.30 and 5.31) showed that there was no contamination beyond
that which would be expected from the preparative technique used to observe the
films. Whilst further study is certainly needed in this area, these results do appear
to confirm that the materials can be fabricated as LB films of some structural
integrity, and suggest that the films can be prepared with some degree of
confidence.
5.4 General Discussion
A detailed study of Langmuir (ie those films on the water surface) and Langmuir-
Blodgett films of Zwitterionic adducts of TCNQ - general formula R(4)Q3CNQ,
where R is an alkyl chain of 8 carbon atoms or greater - has been carried out. The
study concentrated on quinolinium adducts rather than picolinium adducts because it
was felt that the extra aromatic ring of the quinolinium part of the 'head' group
would enable shorter chain adducts to be fabricated as LB films. The ability to
fabricate LB films of anthracene with hydrophobic 'tail' groups as short as 4 carbon
atoms was first shown by workers at Durham University and ICI in the late 1970's
and was discussed in Chapter 3. The structural formula of anthracene is that of 3
fused benzene rings, and this factor was proposed as being the reason why shorter
164
Fig 5-27 : Film Morphology of an 8-layerLB f i l m o f C 1 6 ( 4 ) Q3CNQ
JbS
•> 4 a t i
Fig 5.28 : Diffraction Patterns of an 8-layer
LB f i l m o f C 1 6 ( 4 ) Q3CNQ
I
Fig 5.29 : Diffraction Pattern of an 8-layer
LB f i l m o f C 1 6 ( 4 ) Q3CNQ
<o1
27 CNT IK FS: P10200 EV 20 EV/CHPN
KLM MPRKERS FOR 2 = 7 3 CPU)
Figure 5.30 X-ray spectra of films of C16(4)Q3CNQ showing no contamination beyondwhich would be expected
168
26 CNT IK FS: A10200 EV 20 EV/CHAN
0.0 ;20.3
Figure 5.31 Further X-ray spectra of C16(4)Q3CNQ LB Films
169
hydrophobic tails than those commonly used could be incorporated.
Before discussing in detail mechanisms that may be occurring in the films as the alkyl chain
length is increased, a summary of the results obtained is given.
(i) For alkyl chains <15 carbon atoms, the area per molecule of R(4)Q3CNQ adducts on
the water surface is 32 ±2 A 2.
(ii) . For alkyl chains >15 carbon atoms, the area per molecule of R(4)Q3CNQ adducts is
44 ± 4 A 2.
(iii) For alkyl chains <15 carbon atoms, the resultant LB films are green, with 613 ± 3 nm.
(iv) For alkyl chains >15 carbon atoms, the resultant LB films are purple with 565 ± 4 nm.
In addition, the shorter chain adducts show a relatively broader transition (HWHM = 37 ± 2
nm) compared to the longer chain adducts HWHM = 22 ± 1 nm, in this case).
Clearly, on the basis of the differing area per molecule values, it can be postulated that a
change in molecular orientation or alignment occurs on the water subphase, once the alkyl
chain length is > 15 carbon atoms.
Fig 5.32 shows the possible ways in which the limiting areas of these materials
may alter on compression. Fig 5.32(a) and 5.32(b) show that the limiting
cross-sectional area 'A' can be reduced or increased by reducing or increasing the
degree of overlap between the chroniophores. Fig 5.32(c) shows an alternative where the
170
0
< -A
9 a ^ 6 b > Be
Figure 5.32 Representation of the ways in which the area per molecule, 'A1, can changt
171
angle of tilt of the chromophore, 0 , is reduced, though the degree of overlap
remains the same.
Examination of figures 5.32 and 5.33 shows that perhaps the most reasonable
explanation is that there is a decrease in the value of & so that this leads to a more
favourable electrostatic orientation. Implicit in this model is the requirement that
the cationic portion of the chromophore lies directly above the anionic portion of
the adjacent molecule. This assumption is made purely on the basis of the fact that
it is in this orientation that repulsive electrostatic interactions are minimised.
It is obvious from the'values of the area per molecule that the chromophore is tilted.
The face area is about 144& and the area of the terminal -C(CN)2 "butterfly" is about
24X. As the measured values of the area per molecule lies between these two
extreme positions, then the chromophore must be tilted to some degree. The value
of ©, and thus the orientation of the chromophore, can be calculated as shown
below.
The tilt angle in question, 0 , is the angle between the plane of the subphase and the
long axis of the chromophore. Assuming that the molecules are orientated as shown
in figure 5.34, with 0 in this case purely arbitrary, the parameters needed to
calculate © are:
a) the length (long axis) of the chromophore;
b) the width of the chromophore on its widest part;
c) the face area of the chromophore.
Therefore, from fig 5.34, the length of the chromophore is 'm ', the width at its
widest part is V which can be calculated know ing'm ' and the face area. If the
area per molecule (A) is known, then T can be calculated as 1 = A/n. Then the
172
Figure 5.33 Electrostatic interactions on altering the tilt angle 0
ie: as 0 decreases, so attractive forces increase and repulsive force* decrease
173
C H R O K O P H O R E
Figure 5.34 Schematic diagram showing the parameters needed to calculate 0
174
value of 0 is found as cos & — */ . For the range of adducts studied, the values of
are shown in table 5.5.
The cross-sectional area occupied by a long hydrocarbon chain is about 19X.4 For
confirmational circumstances it is preferable that the long chains align themselves
vertically and parallel as at the orientation the "zig-zag" chains interlock exactly.
For area per molecule values greater that 19X then the chains will tilt to try and
adopt the preferred orientation. Using the model shown in figure 5.35, knowing the
area per molecule of both the chains and the chromophore, the tilt angle of the
chains, <P , can be calculated. The values are shown in table 5.6.
175
R ( n o . o f c a r b o n a t o m s ) 0 *
C 8 7 3
c9 7 5
C I O 7 2
C l l 7 5
C 1 2 7 5
C 1 3 7 4
C 1 4 7 4
C 1 5 6 9
C 1 6 6 7
C 1 8 6 8
C 2 0 6 5
Table 5.5 The variation in the tilt of the R(4)Q3CNQ chromophore with changein hydrophobic chain length
176
CH?CK»C?'ORE
Figure 5.35 Schematic diagram showing the parameters needed to calculate cj)
’a' = area occupied by closely packed alkyl chains 'b' = area per molecule on water subphase (p = resultant tilt angle of the alkyl chains
177
R ( n o . o f c a r b o n a t o m s <£> 0
C 8 3 6
C 9 3 9
C I O 3 4
C l l 3 9
C 1 2 3 9
C 1 3 3 8
C 1 4 3 8
C l 5 2 8
C 1 6 2 5
C 1 8 2 7
C 2 0 2 3
Table 5.6 The variation in the tilt of the hydrocarbon chains as the chain length increases for R(4)Q3CNQ adducts
178
Electrostatic interactions have been postulated a number of tim es5 as the main (or
at least one) factor governing the orientation of zwitterionic amphipathic molecules
at the air-water interface. It is known that an array of parallel dipoles orientated
perpendicular to the water subphase would induce a repulsive interaction. The
alternative attractive interactions would occur when the molecular dipoles are
aligned parallel. The sign and magnitude of any resultant interaction is governed by
the relative orientation of the molecule.
Thus, as a first approximation, it is reasonable to assume that the molecules are
orientated as shown in the model, thus minimising repulsive electrostatic
interactions. The validity of this approximation could be checked by modelling the
monolayer by using energy minimisation techniques.6 The effects of these
electrostatic interactions and other dispersion interactions within the molecule could
then be studied for the range of adducts of increasing alkyl chain length. Other
techniques could also be used to probe the changes occuring in the subphase
surface. For example, grazing - angle fluoresence spectroscopy7 could give
information on the molecular orientation at the air-water interface, or indeed any
aggregation of the molecules that may or may not be present. It is very likely that
should aggregation be occuring, then changes in the area per molecule would be
observed. Thus, an alternative explanation to the observed phenomena involves the
preferential organisation of the molecules into dimers/trimers/tetramers, etc. This
proposal is expanded on in more, detail later in this chapter.
What is clearly very interesting in this study is the observed change in film
behaviour that occurs as the alkyl chain length is increased. It is also noted that this
change is abrupt and therefore cannot be put down to gradual orientational changes
due to subtle hydophobic\hydrophilic interactions. Very little work in this field
systematically looks at altering the alkyl chain length for a common head group, and
observing the effect that this has on film formation - both on the water subphase and
179
solid supports. Data of this type was carried out on fatty acid derivatives early in the century
and is documented in the classic text by Adam4. It appears not to be done to any great extent
these days though this work appears to show that there is much use in carrying out such
studies whenever a new molecule is investigated. Such information can then be used to decide
on the optimum conditions for films formation. Notice that in doing this, we are including
molecular geometry as well as more generally acknowledged factors such as compression
rates, pH, etc in our analysis.
To recap therefore, for C<15, the area per molecule values are smaller than for 0 1 5 . This is
certainly a true phenomenon as can be seen by the X max (nm) values of the resultant LB films
shown previously at table 5.2. There is a clear shift in the characteristic absorption bands.
Notice also that the absorption line is broader for short chains than for long.
A number or reasons can be proposed as to why there is also a change in X max (nm) of the
films with changing alkyl chain length.
The first point that can be made is that the colour of the compressed films on the water
subphase is the same as the colour of the resultant LB films. This can be taken as
evidence that the structural orientation of the Langmuir films is maintained through to
the LB films. If this is the case, then one possible explanation is that the electronic
transition changes from, primarily, an intermolecular to an intramolecular charge transfer
process, as follows. For shorter chain analogues (R<15 carbon atoms) then the chromophore
is tilted more away from the plane of the subphase; for longer chain analogues, (R > 15 carbon
atoms) the chromophore is tilted more towards the subphase plane. This same alignment is
then maintained in the LB film. Thus, for short adducts, the intramolecular charge transfer
transition moment and the electric field vector are aligned in a more parallel arrangement, and
180
thus the resulting transition is primarily an intermolecular process. For long chain
adducts the converse is the case. Here, the intermolecular charge transfer transition
moment and the electric vector are aligned more parallel and hence the resultant
transition is primarily an intramolecular process.
The assignment of the transitions in this way is consistent with the work of Akhtar
et al8 (see Chapter 1) who assigned the lower energy transition, ie longer
wavelength, to an intermolecular transition in their analysis of the solid state spectra
of adducts of this type. Although the energy of the transition in the LB films is
considerably shifted from that of the single crystal spectra they studied, this may
well be due to preferential alignment in the crystal (ie head-to-tail) not wholly
observed in the LB film.
Because the make up of the chromophore is constant throughout the series of
adducts studied during this work, then the changes in area per molecule of adducts
observed during the study, ^ max and shape of the absorption line are probably a
function of changing the alkyl chain length. The change in the parameters occurs at
an alkyl chain length of fifteen carbon atoms, and it may be significant that the
length of a fifteen carbon alkyl chain is the same as the long axis of the TCNQ
quinolinium chromophore. Thus, if it is assumed that there is a subtle interplay
between electrostatic and amphipathic balances, then for alkyl chains less than 15
carbon atoms the hydrophilic moiety is longer than the hydrophobic moiety and for
chains greater than 15 carbon atoms the hydrophilic portion is shorter.
It is also possible that steric factors cause the shifts in area per molecule and
corresponding change in spectra, for example, shorter chain analogues (<C 15),
adopting the orientational characteristics of long chain analogues - 0 =a,65 ° and
cj> =<^23° - would experience some degree of steric hindrance as shown in
figure 5.36.
181
» ** -
Thus, because the colour of LB films is the same as the colour of the compressed Langmuir
film on the surface, it is proposed that the colour arises because the same molecular
orientation is present in both films. Spectroscopic data has already been shown in table 5.4
and for full comparison the collected data in Table 5.7 shows the dramatic change in the
phenomena with increasing chain length. Hence the observed change in the values of 0, c|), X
max (nm) and Area per molecule (A2) in LB and Langmuir films respectively and the shift and
change in shape of the absorption band must be a consequence of this observed change in the
film structure. Comparison with the solution spectra shows that the predominant band is
considerably shifted to shorter wavelength and is much narrower. On the basis of the more
random nature of solution compared to the relatively ordered regime of the LB film, it would
be expected that the transition would be considerably altered.
However, an alternative explanation to those already proposed, and briefly mentioned earlier is
that there is some form of aggregation taking place between molecules. It is very well
documented that many classes of dyes exhibit colour changes due to environmental effects9.
These colour changes and spectral shifts can be split into two categories:
(a) colour shifts due to isomerism
(b) colour shifts due to changes in the environment.
Colour shifts due to isomerism usually involve two distinct species each with
their own particular absorption wavelength. Colour shifts due to changes in the
environment are usually applicable to a single chromophore, and involve steric
interactions, solvent induced shifts or the influence of neighbouring molecules-the well
documented H-and-aggregation phenomena.10 If indeed any aggregation is
occurring, then the shifts observed in this work are undoubtedly environment induced,
183
R ( n o . o f c a r b o n a t o m s ) A . M o l - 1 ( & 2 ) m a x ( n m )
C 8 7 3 3 2 6 1 6
C 9 7 5 3 0 6 1 7
C I O 7 2 3 4 6 1 6
C l l 7 5 30 6 1 6
C 1 2 75 3 0 6 1 3
C 1 3 7 4 3 1 6 1 5
C 1 4 7 4 3 1 6 1 5
C 1 5 6 9 4 0 5 6 1
C 1 6 6 7 4 4 5 6 5
C 1 8 6 8 4 2 5 6 5
C 2 0 6 5 4 8 5 6 8
HWHM
= 3 7
± 2 n m
HWHM
= 22 ± l n m
Table 5.7 Change in parameters with increasing chain length for R(4)Q3CNQ adducts
184
and are probably due to the influence of neighbouring molecules. Indeed, the
classic documentation of steric factors influencing spectroscopic transitions was a
study of quinolinium type dyes.11 It is unlikely however that steric factors alone are
causing the shifts shown in this work - 1,4 - substitution is the least likely
substitution to lead to steric hindrance. Certainly it would be expected that long
chain substitution in the 1 ,2 - position would have a greater effect.
Interactions between neighbouring dye molecules produce spectral shifts and very
definite changes in the band shape. The absorption maxima can shift to shorter
wavelength (H - aggregation) or to longer wavelength (J -aggregation).
As well as generally being identifiable from the changes in the energy of the charge
transfer band, other features are common to H or J aggregates. For example, H
aggregates are favoured by an increase in the alkyl chain length12. This is
interesting when one considers the shifts observed in this work. The longer chain
adducts show an abrupt shift to shorter wavelength at R >C15. It is characteristic
of H and J aggregation that the observed shifts occur abruptly. Thus, there is
evidence, albeit tentative, from the results obtained here for H aggregation - where
longer chains show a shift to shorter wavelength. While generally H aggregated
bands are broad13, there are examples of very sharp bands of this type12.
For dipolar molecules such as the ones studied in this work, the position of allowed
electronic transition is a function of the angle a as shown in figure 5.37.
The angle a is that between the transition moment - in this case the long axis of the •
zwitterion - and a line joining the centres of the molecules. Dipole - dipole
interactions suggest that the lowest energy transition will be allowed for a — > 0 °
and the higher energy transition for a —>90°. At intermediate angles and less ideal
dipole arrangements, more than one transition may occur and the band may be split
Figure 5.37 Allowed electronic transitions as a funtion of the angle 'or'
M
E r 8
0 = 0*
186
- the so called Davydov splitting. It should therefore be expected that the transition
is considerably shifted to higher energy (ie shorter wavelength) in the LB film of
zwitterionic adducts in view of the calculated values of the tilt angle assuming that
the LB films are produced with a certain degree of structure.
This discussion has highlighted possible explanations for the observed changes in
film behaviour with increasing chain length. However, further work is needed to
fully ellucidate the nature of these changes. The possibilities for further work are
many, though all would involve more intensive investigation of film structure. For
example, low-angle x-ray diffraction would enable accurate determination of the
interlayer spacings within the LB film. Such information would enable us to
conclude whether or not the tilt angles calculated in this work are accurate. It
would also give data on the specific orientation of the molecules - though this can
be determined more accurately from the polarised uv-vis or ir-spectra of the
molecules. Very simply, as the spectra are determined by interaction of a
molecular transition moment with the electric field vector of the incident radiation,
then systematic orientation of the LB films, at angles to the incident radiation, can
give by analysis of the spectra, accurate determination of the molecular orientation.
For example, if the electric field vector and the molecular transition dipole are
orthogonal, then no, or very weak, spectral bands will be observed.
Thus, a combination of microscopy, diffraction and spectroscopy could probe the
structural changes taking place. Because, such a change in the film structure leads
to quite large changes in the resultant uv spectra, then for some potential
applications this could be a possible advantage of these materials over others. For •
example, for photochromic devices, mixed monolayers could be produced between
long chain and short chain adducts with the resultant overlap in the uv spectra
giving a tunable response compared to simple monolayers of only long or short
chains. Similarly, such tunable responses would be useful for any optical second
187
harmonic generation applications.
There is much scope for the synthesis of derivatives of this one general type of
zwitterionic adduct. Long chain adducts of -rather than 'jf -substituted quinoline
could be attempted, as could o( - or ^ -substituted picoline adducts. There is also
scope for fluorination of the alkyl chains as well as using derivatives of TCNQ, eg
TCNQF4. The extent to which adducts of these types show similar changes in film
behaviour over a homologous series could then be studied, as well as any
advantages/disadvantages in film stability.
However, to conclude this portion of the discussion therefore, it is postulated that
the structure of the films is governed primarily by electrostatic interactions within
the head groups. The chromophore is orientated intermediate between an
arrangement parallel and perpendicular to the subphase and support. As the
hydrophobic nature of the material increases, the molecule undergoes another
rearrangement. The hydrophobic chains do not lie perpendicular to the plane of the
film, rather they are tilted towards the plane. The tilt angles of the chromophore, 0 1
and the hydrocarbon chain, cj) , are slightly different which results in a slight
herringbone arrangement. The proposed film structure is shown in fig 5.38. The
absorption band of this film is shifted considerably to shorter wavelength possibly
due to aggregation and the dipole - dipole interactions resulting between neighbours,
or a change from an inter- to an intra-molecular charge transfer. These tentative
explanations are worthy of further work in this area.
5.5 Change in Film structure and absorbance bands with time
The area per molecule values quoted are obtained when the film is deposited and
compressed immediately. In this case, "immediately" implies as soon as the usual
period of time has elapsed between deposition of material onto the subphase and
188
Figure 5.38 Proposed structure of the fabricated LB film of long chain analogues of the zwitterion R(4)Q3CNQ - particular tilt angles applicable to C11(4)Q3CNQ
189
then allowing time for the material to evaporate. If, however, the film is deposited
and then left for increasing periods of time, then the area per molecule can increase
dramatically. This is shown for various adducts in table 5.8. The resultant shifts
occurring in the spectra of the LB film deposited after these time periods is shown
in fig 5.39. From the previous discussion, it could be concluded that the molecules
are tilted further and further towards the subphase, and preserving this structure
within the LB film as a consequence. Thus, the tilt angle 0 decreases to a value of
^28° with a corresponding decrease in the tilt of the hydrocarbon chains. Such a
tilt angle ( a -> 0 ) , from the previous argument) would be expected to lead to shifts
to longer wavelengths.
However, it is known that these materials are photochromic and a one other
plausible explanation is that the gross reorientation that appears to be occurring at
the subphase is due to a photochromic reaction causing gross molecular
rearrangement within the chromophore. Alternatively, further aggregation of the
H- or J-types in films is occurring - it is clear that further work is needed in this
area also.
190
Figure 5.39 Change in UV/VIS spectra with time for C16(4)Q3CNQ
1 Compressed immediately
2 Compressed after 2 V i hours
see400
0
191
A D D U C T i 0A R E A P E R M O L E C U L E 1 ( A 2 )
C l l ( 4 ) Q 3 C N Q
I m m e d i a t e 2 6
2 ^ h o u r s 4 5
C 1 5 ( 4 ) Q 3 C N Q
I m m e d i a t e 4 0
2 h h o u r s 6 2
Thble 5.8 Change in Area Per Molecule with time for C l l (4) Q3CNQ and C15 (4) Q3CNQ
192
5.6 REFERENCES
1 G L Gaines, "Insoluble Monolayers at Liquid - Gas Interfaces", New York, Interscience, (1966)
2 H P Zingsheim, Scanning Electron Microsc., I , 357, (1977)
3 See for example, R A Hann, S K Gupta, J R Fryer and B L Eyers, Thin Solid Films, 134, 35, (1985)
4 N K Adam, "The Physics and Chemistry of Surfaces", Oxford University Press
5 P J Bowen and T J Lewis, Thin Solid Films, 99, 157, (1983)D A Cadenhead, R J Demchalk and M C Phillips, Kolloid-ZZ-fiir Polymere, 220. 63, (1967)
6 A Ulman, S D Evans and R G Snyder, Thin Solid Films, 210, 806, (1992)
7 R Linton, V Guarisco, J J Lee, B Hagenhoff and A Benninghoven, Thin Solid Films, 210, 565, (1992)
8 S Akhtar, J Thnaka, R M Metzger and G J Ashwell, Mol Cryst Liq Cryst, 139. 353, (1986)
9 The Theory of the Photographic Process (4 ed), Ed T H James, Macmillan Publishing, New York, (1977)
10 H Bucher and H Kuhn, Chem Phys Lett, 6, 183, (1970)
11 L G S Brooker, F L White, R H Sprague, S G Dent Jnr and G Van Zandt, Chem Rev, 41, 325, (1947)
12 A H Herz, Photograph Sci Eng, 18, 323, (1974)
13 W West, S P Lovell and W Cooper, Photograph Sci Eng, 14, 52, (1970)
193
ACKNOWLEDGEMENTS
I would like to express my warmest thanks to Dr Norman Bell of the Department of
Applied Chemistry and Professor John Brooks of the Materials Research Institute,
Sheffield Hallam University, for all their support and encouragement throughout this
work. I also thank all the staff in the Environment Sensing Systems Section of the
Health and Safety Executive's Research Laboratories in Sheffield - in particular Dr
Stephen Thorpe without whose support much of this work would have been impossible.
Thanks are also due to Dr Marek Szeblewski for many valuable discussions and loan of
materials, and also Dr Ian Sandy for providing me with the materials used in the early
part of the work. I would also like to thank Dr Tom Maclean of ICI and
Drs Graham Cross, Ian Girling and Ian Peterson, late of the long range research group
at the GEC Hirst Laboratories, for advice in establishing the Langmuir-Blodgett film
laboratory and allowing me to spend time at their laboratories. For helpful discussions
on interfacial phenomena I would like to thank Dr Malcolm Macdonald and, for help in
the interpretation of the mass spectra, Dr John Little, both of Sheffield Hallam University.
I would also like to thank all the technical staff at Sheffield Hallam University and the
other postgraduate students from both the Department of Chemistry and Applied Physics;
Mrs Suzanne Wragg for typing the work; my family for putting up with someone who
refused to foresake the life of a student until the age of twenty-five, and Curtise Monk for
her support.
Finally, I would like to pay tribute to the late Dr Alwyn Jones of the Health and Safety
Executive, Sheffield, who shared in the supervision of this work, and whose unerring
support through the often frustrating times of developing the Langmuir-Blodgett
laboratory was invaluable. As well as being a much respected scientist, he was a man of
great warmth, and those who had known and worked with him would, I think, say it was
a pleasure that they had done so.